Você está na página 1de 113

doi:10.

1093/jicru/ndi016
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
Report Committee
J. Zoetelief (chairman), Faculty of Applied Sciences, Delft University of Technology,
Delft, Netherlands
D. R. Dance, The Royal Marsden NHS Foundation Trust, London, United Kingdom
G. Drexler, GSF-National Research Center for Environment and Health, Neuherberg, Germany and
State University of Rio de Janeiro, Rio de Janeiro, Brazil
H. Ja rvinen, Radiation and Nuclear Safety Authority (STUK), Helsinki, Finland
M. Rosenstein, Clarksburg, Maryland, USA
Commission Sponsors
H. G. Paretzke, GSF-National Research Center for Environment and Health, Neuherberg, Germany
K. Doi, The University of Chicago, Chicago, Illinois, USA
A. Wambersie, Universite Catholique de Louvain, Brussels, Belgium
Consultants to the Report Committee
P. Allisy-Roberts, Bureau International des Poids et Mesures, Sevres, France
H. Bosmans, University Hospital Gasthuisberg, Leuven, Belgium
C. J. Moretti, National Physical Laboratory, Teddington, UK
J. Van Dam, University Hospital Gasthuisberg, Leuven, Belgium
E. Va~nnoo, Complutense University, Madrid, Spain
B. F. Wall, Radiation Protection Division, Health Protection Agency, Chilton, UK
The Commission wishes to express its appreciation to the individuals involved in the preparation of this report,
for the time and efforts which they devoted to this task and to express its appreciation to the organizations with
which they are affiliated.
All rights reserved. No part of this book may be reproduced, stored in retrieval systems or transmitted in any
form by any means, electronic, electrostatic, magnetic, mechanical photocopying, recording or otherwise,
without the permission in writing from the publishers.
British Library Cataloguing in Publication Data. A Catalogue record of this book is available at the British
Library.
ISBN 0199203208
Journal of the ICRU Vol 5 No 2 (2005) Report 74 doi:10.1093/jicru/ndi017
Oxford University Press
THE INTERNATIONAL COMMISSION ON RADIATION UNITS
AND MEASUREMENTS
INTRODUCTION
The International Commission on Radiation Units
and Measurements (ICRU), since its inception in
1925, has had as its principal objective the develop-
ment of internationally acceptable recommenda-
tions regarding:
(1) quantities and units of radiation and radio-
activity,
(2) procedures suitable for the measurement and
application of these quantities in clinical radio-
logy and radiobiology, and
(3) physical data needed in the application of these
procedures, the use of which tends to assure
uniformity in reporting.
The Commission also considers and makes similar
types of recommendations for the radiation protec-
tion field. In this connection, its work is carried out
in close cooperation with the International Commis-
sion on Radiological Protection (ICRP).
POLICY
The ICRU endeavors to collect and evaluate the
latest data and information pertinent to the pro-
blems of radiation measurement and dosimetry and
to recommend the most acceptable values and tech-
niques for current use.
The Commissions recommendations are kept
under continual review in order to keep abreast of
the rapidly expanding uses of radiation.
The ICRU feels that it is the responsibility of
national organizations to introduce their own
detailed technical procedures for the development
and maintenance of standards. However, it urges
that all countries adhere as closely as possible to
the internationally recommended basic concepts of
radiation quantities and units.
The Commission feels that its responsibility lies in
developing a system of quantities and units having
the widest possible range of applicability. Situations
may arise from time to time when an expedient
solution of a current problem may seem advis-
able. Generally speaking, however, the Commission
feels that action based on expediency is inadvisable
from a long-term viewpoint; it endeavors to base
its decisions on the long-range advantages to be
expected.
The ICRU invites and welcomes constructive com-
ments and suggestions regarding its recommenda-
tions and reports. These may be transmitted to the
Chairman.
CURRENT PROGRAM
The Commission recognizes its obligation to pro-
vide guidance and recommendations in the areas of
radiation therapy, radiation protection, and the com-
pilation of data important to these fields, and to
scientific research and industrial applications of
radiation. Increasingly, the Commission is focusing
on the problems of protection of the patient and
evaluation of image quality in diagnostic radiology.
These activities do not diminish the ICRUs commit-
ment to the provision of a rigorously defined set of
quantities and units useful in a very broad range of
scientific endeavors.
The Commission is currently engaged in the
formulation of ICRU reports treating the following
subjects:
Approaches to the Dosimetry of Low-Dose Exposures to
Ionizing Radiation
Assessment of Image Quality in Nuclear Medicine
Bone Densitometry
Doses for Cosmic Ray Exposure for Aircrew
Dose and Volume Specifications for Reporting Intracavi-
tary Therapy in Gynecology
Dosimetry Systems for Radiation Protection
Elastic Scattering of Electrons and Positrons
Image Quality and Patient Exposure in CT
Mammography------Assessment of Image Quality
Measurement Quality Assurance for Ionizing
Radiation
Prescribing, Recording, and Reporting Conformal Photon
Beam Therapy
Prescribing, Recording, and Reporting Proton Beam
Therapy
Requirements for Radiological Sampling
ROC Analysis
International Commission on Radiation Units and Measurements 2005
In addition, the ICRU is evaluating the possibi-
lity of expanding its program to encompass non-
ionizing radiation, particularly the quantities and
units aspects.
The Commission continually reviews radiation
science with the aim of identifying areas where the
development of guidance and recommendations can
make an important contribution.
THE ICRUS RELATIONSHIP WITH OTHER
ORGANIZATIONS
In addition to its close relationship with the ICRP,
the ICRU has developed relationships with other
organizations interested in the problems of radiation
quantities, units, and measurements. Since 1955,
the ICRU has had an official relationship with the
World Health Organization (WHO), whereby the
ICRU is looked to for primary guidance in matters
of radiation units and measurements and, in turn,
the WHO assists in the worldwide dissemination of
the Commissions recommendations. In 1960, the
ICRU entered into consultative status with the
International Atomic Energy Agency (IAEA). The
Commission has a formal relationship with the Uni-
ted Nations Scientific Committee on the Effects of
Atomic Radiation (UNSCEAR), whereby ICRU
observers are invited to attend annual UNSCEAR
meetings. The Commission and the International
Organization for Standardization (ISO) informally
exchange notifications of meetings, and the ICRU
is formally designated for liaison with two of the
ISO technical committees. The ICRU also corres-
ponds and exchanges final reports with the following
organizations:
Bureau International de Metrologie Legale
Bureau International des Poids et Mesures
European Commission
Council for International Organizations of Medical
Sciences
Food and Agriculture Organization of the United Nations
International Committee of Photobiology
International Council of Scientific Unions
International Electrotechnical Commission
International Labor Office
International Organization for Medical Physics
International Radiation Protection Association
International Union of Pure and Applied Physics
United Nations Educational, Scientific and Cultural
Organization
The Commission has found its relationship with
all of these organizations fruitful and of substantial
benefit to the ICRU program. Relations with these
other international bodies do not affect the basic
affiliation of the ICRU with the International
Society of Radiology.
OPERATING FUNDS
In recent years, principal financial support has
been provided by the European Commission, the
National Cancer Institute of the U.S. Department
of Health and Human Services and the International
Atomic Energy Agency. In addition, during the last
10 years, financial support has been received from
the following organizations:
Belgian Nuclear Research Centre
Canadian Nuclear Safety Commission
Eastman Kodak Company
Electricite de France
Fuji Medical Systems
Hitachi, Ltd.
International Radiation Protection Association
International Society of Radiology
Ion Beam Applications
Italian Radiological Association
Japan Industries Association of Radiological Systems
Japanese Society of Radiological Technology
MDS Nordion
National Institute of Standards and Technology
Nederlandse Vereniging voor Radiologie
Philips Medical Systems, Incorporated
Radiation Research Society
Siemens
Varian
In addition to the direct monetary support pro-
vided by these organizations, many organizations
provide indirect support for the Commissions pro-
gram. This support is provided in many forms,
including, among others, subsidies for (1) the time
of individuals participating in ICRU activities,
(2) travel costs involved in ICRU meetings, and (3)
meeting facilities and services.
In recognition of the fact that its work is made
possible by the generous support provided by all of
the organizations supporting its program, the Com-
mission expresses its deep appreciation.
Andre Wambersie
Chairman, ICRU
Brussels, Belgium
PATIENT DOSIMETRY FOR X-RAYS USED IN MEDICAL IMAGING
Journal of the ICRU Vol 5 No 2 (2005) Report 74 doi:10.1093/jicru/ndi018
Oxford University Press
PATIENT DOSIMETRY FOR X RAYS USED IN
MEDICAL IMAGING
CONTENTS
PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
GLOSSARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
EXECUTIVE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.1 Evolution of radiation dosimetry in medical x-ray imaging . . . . . . . . . . . . . . . 9
1.2 Risks for the patient in radiological imaging and relevant
dosimetric quantities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.2.1 Acute deterministic effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.2.2 Late effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.2.2.1 Cancer induction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.2.2.2 Late effects in normal tissues . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.2.2.3 Impairment of mental development . . . . . . . . . . . . . . . . . . . . . . . 12
1.2.2.4 Genetic risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.2.3 Relevant dosimetric quantities and dosimetric procedures . . . . . . . . . . . . . . . 12
1.2.4 Required accuracy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.3 Dosimetry in radiology: relevant quantities . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.3.1 Calibration at the Standards Laboratory . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.3.2 From air kerma free-in-air to absorbed dose in water in patient or phantom . . . . 13
1.3.3 Air kerma-area product (KAP) and dose----area product (DAP) . . . . . . . . . . . . . 14
1.3.4 Reporting patient irradiation in radiological imaging . . . . . . . . . . . . . . . . . 14
1.3.4.1 Radiological parameters of the exposure . . . . . . . . . . . . . . . . . . . . 14
1.3.4.2 Air kerma----area product (KAP) or dose----area product (DAP) . . . . . . . . 15
1.3.4.3 Monte Carlo computation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1.3.4.4 Phantoms and in vivo measurements. . . . . . . . . . . . . . . . . . . . . . . 15
1.3.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.4 Need for harmonization of quantities and terminology . . . . . . . . . . . . . . . . . 17
1.5 The two purposes of patient dosimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.5.1 To set and check standards of good practice . . . . . . . . . . . . . . . . . . . . . . . 18
1.5.2 To assist in assessing detriment or harm . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.6 Relationship between patient dose and image quality . . . . . . . . . . . . . . . . . . 18
1.7 Scope of the report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2 SPECIFICATION OF X-RAY BEAMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.1 Photon spectrum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.2 Half-value layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.3 X-ray tube voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.4 Total filtration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.5 X-ray tube output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3 QUANTITIES AND UNITS FOR MEASUREMENT AND CALCULATION IN MEDICAL
X-RAY IMAGING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.1 Basic dosimetric quantities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.2 Application-specific quantities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.2.1 Incident air kerma and incident air kerma rate . . . . . . . . . . . . . . . . . . . . . 28
3.2.2 Entrance-surface air kerma and entrance-surface air kerma rate . . . . . . . . . . . 29
3.2.3 Air kerma----area product and air kerma----area product rate . . . . . . . . . . . . . . 29
3.2.4 Air kerma----length product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.2.5 CT air-kerma index free-in-air . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.2.6 CT air-kerma index in the standard CT dosimetry phantoms . . . . . . . . . . . . . 30
3.2.7 Weighted CT air-kerma index and normalized weighted CT air-kerma index . . . 30
3.2.8 CT air kerma-length product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.3 Risk-related quantities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.3.1 Absorbed dose in relation to deterministic effects . . . . . . . . . . . . . . . . . . . . 31
3.3.2 Absorbed dose for assessment of stochastic effects (organ dose) . . . . . . . . . . . . 31
3.3.3 Equivalent dose and effective dose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.4 Dose-conversion coefficients for assessment of organ and tissue doses . . . . . . . 32
3.5 Quantities recommended for establishment and use of diagnostic
reference levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.5.1 Incident air kerma and entrance-surface air kerma . . . . . . . . . . . . . . . . . . . 34
3.5.1.1 Mean mammary glandular dose . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.5.2 Incident air kerma rate and entrance-surface air kerma rate . . . . . . . . . . . . . 34
3.5.3 Air kerma----area product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.5.4 CT Air kerma----length product, P
DL,CT
. . . . . . . . . . . . . . . . . . . . . . . . . . . 34
4 MEASUREMENT METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.1 Quality assurance of dosimeters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.1.1 Calibration of dosimeters in terms of air kerma free-in-air . . . . . . . . . . . . . . . 36
4.1.2 Calibration of air kerma----area product meters . . . . . . . . . . . . . . . . . . . . . . 37
4.1.3 Calibration of thermoluminescent dosimeters . . . . . . . . . . . . . . . . . . . . . . 38
4.2 Measurement methods for specific dosimetric quantities . . . . . . . . . . . . . . . . 39
4.2.1 Dosimeters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.2.1.1 Ionization chamber dosimeters . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.2.1.2 Thermoluminescent dosimeters . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.2.1.3 Scintillation dosimeters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.2.1.4 Film dosimeters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
4.2.2 Incident air kerma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
4.2.3 Entrance-surface air kerma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.2.4 Air kerma----area product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.2.5 CT air-kerma index and CT air-kerma index in the standard CT head
and body dosimetry phantoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.2.5.1 Pencil ionization chamber dosimeter . . . . . . . . . . . . . . . . . . . . . . 44
4.2.5.2 Thermoluminescent dosimeters . . . . . . . . . . . . . . . . . . . . . . . . . 47
4.3 Features of measurements on patients and measurements with
physical phantoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
4.4 Skin dose determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
4.4.1 Direct measurement of the maximum skin dose . . . . . . . . . . . . . . . . . . . . . 48
4.4.1.1 Skin dose measurements on patients with thermoluminescent
dosimeters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
4.4.1.2 Skin dose measurements on patients with scinillation dosimeters . . . . . 50
4.4.1.3 Skin dose measurements on patients with film dosimeters . . . . . . . . . 50
4.4.2 Derivation of the skin dose from the air kerma----area product P
KA
. . . . . . . . . . 50
4.4.3 Derivation of the skin dose directly from the radiological parameters
of the exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5 METHODS FOR DETERMINING ORGAN AND TISSUE DOSES . . . . . . . . . . . . . . . . . . . . . . . 55
5.1 Dose measurements in physical phantoms . . . . . . . . . . . . . . . . . . . . . . . . . . 55
5.2 Monte Carlo radiation transport calculations . . . . . . . . . . . . . . . . . . . . . . . 56
5.2.1 Main features of the Monte Carlo technique . . . . . . . . . . . . . . . . . . . . . . . 56
5.2.2 Main features of the computational models of the human body . . . . . . . . . . . . 56
5.2.2.1 Mathematical phantoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
5.2.2.2 Special features of the active bone marrow . . . . . . . . . . . . . . . . . . . 57
5.2.2.3 Voxel phantoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
5.2.3 Uncertainties in Monte Carlo organ-dose calculations . . . . . . . . . . . . . . . . . 58
5.2.4 Comparison of conversion coefficients calculated at different institutes . . . . . . . 59
5.2.5 Comparison of measured and calculated organ doses . . . . . . . . . . . . . . . . . . 59
5.2.5.1 Adult phantoms: organs in the x-ray field . . . . . . . . . . . . . . . . . . . 59
5.2.5.2 Adult phantoms: organs outside the x-ray field . . . . . . . . . . . . . . . . 60
5.2.5.3 Adult phantoms: active bone marrow . . . . . . . . . . . . . . . . . . . . . . 60
5.2.5.4 Paediatric phantoms: head and neck . . . . . . . . . . . . . . . . . . . . . . 60
5.2.5.5 Paediatric phantoms: whole body . . . . . . . . . . . . . . . . . . . . . . . . 60
5.2.5.6 Adult phantoms: CT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
5.2.6 Sources of data on dose-conversion coefficients . . . . . . . . . . . . . . . . . . . . . . 61
6 CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
APPENDIX A BACKSCATTER FACTORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
APPENDIX B HANDBOOKS PRODUCED BY THE CENTER FOR DEVICES
AND RADIOLOGICAL HEALTH (CDRH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
APPENDIX C REPORTS PRODUCED BY THE GERMAN NATIONAL RESEARCH
CENTER FOR ENVIRONMENT AND HEALTH (GSF) . . . . . . . . . . . . . . . . . . . . . 79
APPENDIX D REPORTS PRODUCED BY THE HEALTH PROTECTION AGENCY (HPA) (FORMERLY
NATIONAL RADIOLOGICAL PROTECTION BOARD) (NRPB) . . . . . . . . . . . . . . . . 87
APPENDIX E REVIEW OF MONTE CARLO CALCULATIONS FOR ASSESSMENT OF MEAN
GLANDULAR DOSE IN MAMMOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
APPENDIX F PCXMC ------ A PC-BASED MONTE CARLO PROGRAM FOR CALCULATING
PATIENT DOSES IN MEDICAL X-RAY EXAMINATIONS . . . . . . . . . . . . . . . . . . . 99
REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
CONTENTS
Journal of the ICRU Vol 5 No 2 (2005) Report 74 doi:10.1093/jicru/ndi019
Oxford University Press
PREFACE
The mission and main objective of the Interna-
tional Commission on Radiation Units and Measure-
ments (ICRU) are to develop a coherent system of
radiological quantities and units that is accepted
worldwide and applied in all fields where ionizing
radiation is used. The ICRU also develops recom-
mendations on how to measure radiation-related
quantities to ensure a reliable exchange of results.
In addition, within the framework of this mission,
there is often a need for the definition of new terms
and concepts that could be adopted universally. The
ultimate goal of the ICRU is to improve harmoniza-
tion in the concepts and the methods used to describe
and to report radiation applications, and thereby
facilitate the exchange of information between cen-
tres using radiation in medicine, science, and indus-
try. The foundation of the ICRU by the First
Congress of Radiology in 1925 was to solve exactly
this harmonization problem.
The present Report is the first report published
by the ICRU that deals with patient dosimetry for
x rays used in diagnostic medical imaging. The
impetus for this report derives from the broad and
systemic application of x rays for diagnostic and
interventional imaging. The increasing number of
patients that benefits from radiology and the
increasing number and types of procedures that are
applied to these patients have resulted in a dramatic
increase of the population dose, which, in developed
countries, often exceeds the natural radiation levels.
The situation in developing countries will sooner or
later exhibit the same trend (UNSCEAR, 2000).
The relation between image quality and patient
dose has always been a matter of concern for the
radiology community. To initiate the production of
objective information, the European Commission
conducted several trials for various types of exam-
ination currently performed in diagnostic radiology.
As an example, a first trial, involving 24 radiology
departments from 10 European countries
(1987/1988) compared entrance doses for PA chest
radiography. The entrance surface doses ranged
from0.03 to 12 mGy, i.e., a ratio of about 400 (Maccia
et al., 1989). A second larger study (1991) involved
83 radiology departments from 16 countries. As
an example, in the second study, for PA chest radio-
graphs, the mean entrance doses measured in
the participating departments ranged from 0.1 to
0.5 mGy, i.e., a ratio of 5 between the maximum
and the minimum mean doses (EC, 1996a). This
ratio was worse in the first study. Most interesting
is the fact that there was no correlation between the
quality of the image and the dose to the patient. The
quality of the images was evaluated by the informa-
tion content of the film as assessed by a team of
experienced radiologists. Several other studies of
this kind were initiated (e.g., for breast, lumbar
spine), and quality criteria were established (CEC,
1990; EC, 1996b; ICRU, 2003).
A particular source of concern is that among the
different examinations some modern CT procedures
that are remarkably powerful in their diagnostic
capabilities deliver significant doses to large regions
of the body. The diagnostic power of the radiological
procedures to solve medical issues is of course the
first priority. Because of the increasing doses
delivered to an increasing number of patients, how-
ever, it becomes important and timely to optimize
the technical conditions, i.e., to reduce the patient
exposure for the same quality of diagnostic informa-
tion. This is simply common sense and is in agree-
ment with the recommendations of the radiation
protection commissions and agencies, and also of
national and international authorities.
An additional issue is the recent development and
rapid growth of interventional radiology, especially
in cardiology. The exposures are high for the patient
(and possibly also for the radiologists), and the num-
ber of reported cases of acute tissue reactions
1
with
different severity is increasing. The need for accur-
ate dosimetry becomes critical especially for the
skin, which is one of the tissues at highest risk.
For patient dosimetry in radiology, the required
accuracy depends on the clinical situation and the
dose range involved. It is, in general, much lower
than the accuracy required in radiation therapy. In
any case the quantities measured should always be
1
The ICRP has recently proposed (2005) to replace the
term deterministic effects by tissue reactions.
International Commission on Radiation Units and Measurements 2005
clearly identified. The relevant quantities to be
determined are those most closely related to the bio-
logical effects or risks of such effects. Presently, the
available data establish correlation between the bio-
logical effects and absorbed dose at the point or in
the volume of interest.
For the low doses delivered by most of the current
procedures in diagnostic radiology, cancer induction
(stochastic effect) is considered to be the main risk.
In contrast to diagnostic radiology, at the high doses
delivered, for example, during interventional radi-
ology, acute effects become the major source of con-
cern (deterministic effect). Late effects resulting
from acute effects constitute particularly dangerous
pre-cancerous lesions. This is well established for
the skin and is nowbeing investigated for the rectum
after radiotherapy of prostate and cervix tumors.
Previous ICRU Reports have dealt with patient
dosimetry for external beam therapy with photons
(ICRU Report 42 1987; ICRU Report 50, 1993; ICRU
Report 62, 1999; ICRU Report 64, 2001), with elec-
trons (ICRU Report 71, 2004a), protons (ICRU
Report 59, 1998b), and neutrons (ICRU Report 45,
1989a), for brachytherapy, and for b-ray applica-
tions (ICRU 38, 1985a; ICRU Report 58, 1997;
ICRU Report 72, 2004b). ICRU Reports 32 (1979)
and 67 (2002) dealt with patient dosimetry in
diagnostic nuclear medicine procedures.
Some aspects of the dosimetry of x rays generated
at tube voltages ranging from 5 to 150 kV were dis-
cussed in ICRU Report 17 (1970) but were not direc-
ted specifically at patient dosimetry in diagnostic or
interventional radiology.
In the field of radiation protection, the ICRU has
also provided advice on the determination of dose
equivalents from sources of radiation external to
the body (ICRU Report 39, 1985b; ICRU Report 43,
1988; ICRU Report 47, 1992a; ICRU Report 57,
1998a). The advice involved the use of operational
quantities suitable for practical measurements for
the evaluation of occupational exposures. In particu-
lar, these operational quantities introduced by the
ICRU facilitate an adequate and conservative estim-
ate of effective dose, that is, the protection quantity
defined by the International Commission on Radio-
logical Protection (ICRP) for use in its system of
radiological protection (ICRP Publication 60, 1991a).
Exposure to ionizing radiation due to medical
x-ray imaging entails the well-defined irradiation
of localized parts of the body. Diverging beams of
x rays, emitted from a point source and character-
ized by half-value layers of 0.310 mm of Al, are
collimated to penetrate the volume of interest. The
use of different irradiation conditions, in terms of
incident radiation quality and beam geometry in
relation to the patients body, has led to the develop-
ment of specific dosimetric methods and the defini-
tion of appropriate quantities quantities different
from those used for occupational and environmental
exposures.
Not surprisingly, the exposure conditions
assumed in deriving the relationships between the
effective dose and the operational quantities for
occupational and environmental exposures are not
appropriate for patient dosimetry in medical
imaging. In the first case one is dealing with whole
body irradiation by broad beams of photons or neut-
rons while, in the second case, strictly collimated
beams are used resulting in partial-body patient
irradiations.
Whereas some of the dosimetric concepts and tech-
niques used in radiotherapy have been successfully
employed in medical x-ray imaging, additional dosi-
metric quantities and measurement methods are
required for patient dosimetry for procedures such
as fluoroscopy, CT, and mammography. Conversion
coefficients are often used in practice to relate dir-
ectly measurable quantities to doses to different
critical organs or at specific reference points. When
deterministic effects are considered a possibility,
doses to the more heavily irradiated sites of the
body need to be critically evaluated.
The present Report provides a detailed framework
of recommendations for assessing patient dose in
radiological imaging. Moreover, this framework is
suitable for the accurate, harmonized exchange of
information as well as to provide an assessment to
avoid or reduce the severity of tissue reactions. This
report will be soon followed by a second one, exclus-
ively focused upon CT dosimetry and its image qual-
ity. The recent development and rapid growth of CT
applications, and the specific issues that are raised,
deserve a special ICRU Report.
Andre Wambersie
Paul M. DeLuca
Johannes Zoetelief
September 2005
PATIENT DOSIMETRY FOR X RAYS USED IN MEDIUM IMAGING
2
Journal of the ICRU Vol 5 No 2 (2005) Report 74 doi:10.1093/jicru/ndi020
Oxford University Press
GLOSSARY
AEC automatic exposure control
AP anteroposterior view: x rays enter from the front of a patient
BM bone marrow
CC cranio-caudal view: x rays enter on the top of the head of a patient
CT computed tomography
CTDI computed tomography dose index
DICOM digital imaging and communications in medicine
DRL diagnostic reference level
FID focal spot-to-image receptor distance
FOV field of view
FSD focal spot-to-surface distance
GI gastro-intestinal
GSD genetically significant dose
HVL half-value layer
LAT lateral incidence of radiation
LAO left anterior oblique view: x rays enter right rear side of patient and form an image on the left front
side
L LAT left lateral view: x rays enter from the right side of patient and form an image on the left side
LPO left posterior oblique view: x rays enter right front side of patient and form an image on the left rear
side (All oblique views lie in a transverse plane and form a 45

angle with the AP/PA axis.)


LSJ lumbosacral joint
ML medio-lateral
MRI magnetic resonance imaging
OBL oblique
PA posteroanterior view: x rays enter rear of patient
PMMA polymethylmethacrylate plastic
RAO right anterior oblique view: x rays enter left rear side of patient and forman image on the right front
side
R LAT right lateral view: x rays enter left side of patient and form an image on the right side
RPO right posterior oblique view: x rays enter left front side of patient and form an image on the right
rear side
TLD thermoluminescent dosimeter
TMJ temporomandibular joint
International Commission on Radiation Units and Measurements 2005
Journal of the ICRU Vol 5 No 2 (2005) Report 74 doi:10.1093/jicru/ndi021
Oxford University Press
ABSTRACT
This report presents specifications of x-ray beams
and quantities, and units for dose measurement and
calculation in medical x-ray imaging, including
application-specific quantities, and new symbols. It
addresses measurement methods for normalization
quantities and for quantities recommended for the
establishment and use of diagnostic reference levels.
It presents methods for determining organ and tis-
sue doses as well as doses in localized regions of
organs and tissues, including detailed information
on dose-conversion coefficients for x-ray imaging
fields. This is the first ICRU report dealing with
methods for patient dosimetry of x rays used in med-
ical imaging. Whereas some of the dosimetric con-
cepts and techniques used in radiotherapy have been
successfully employed for medical imaging using x
rays, additional dosimetric quantities and measure-
ment methods are required for patient dosimetry
associated with procedures such as radiography, flu-
oroscopy, CT, and mammography.
International Commission on Radiation Units and Measurements 2005
Journal of the ICRU Vol 5 No 2 (2005) Report 74 doi:10.1093/jicru/ndi022
Oxford University Press
EXECUTIVE SUMMARY
In Section 1 it is emphasized that medical
imaging was virtually the first application of
x rays. The health risks of irradiation became
evident only later. Appropriate quantities to meas-
ure the amount of irradiation of an object had to be
developed, leading to quantities like exposure, ab-
sorbed dose, and kerma. It is furthermore stressed
that there is a need for harmonization of quantities
and terminology for different applications in med-
ical imaging using x rays. The two purposes of
patient dosimetry of x rays used in medical imaging
are to set and check standards of good practice and
to assess the risks of detriment or harm. Image
quality is stressed to be of paramount importance
in medical imaging but is the subject of other ICRU
Reports.
Specifications of x-ray beams used for medical
imaging are presented in Section 2. It is recommen-
ded to characterize the radiation quality of x-ray
beams used for medical imaging by a combination
of various parameters, including first and second
half-value layer, HVL
1
and HVL
2
, the ratio of
HVL
1
and HVL
2
, the tube voltage, and the total
filtration. In most cases a combination of three of
these parameters will be sufficient for character-
ization. The radiation intensity is also an import-
ant characteristic of an x-ray tube (including
filtration). For this purpose the x-ray tube output
is defined.
Quantities and units for dose measurement
and calculation in medical x-ray imaging are
dealt with in Section 3. Relevant basic dosimetric
qualities are presented in first instance. Several
application-specific qualities have been found use-
ful for measurements in medical x-ray imaging,
but ambiguity remains in the names of quanti-
ties and their use. Therefore, application-specific
quantities and new symbols are defined. Concern-
ing risk-related quantities, mean organ and tissue
doses are defined as well as absorbed dose to the
more heavily irradiated regions of tissues in rela-
tion to deterministic effects. Dose-conversion coef-
ficients relate the specified dosimetric quantities to
a normalization quality. Both types of dosimetric
quantity are discussed in Section 4. Quantities are
also recommended for establishment and use of
diagnostic reference levels (DRLs). It should be
noted that the recommendations made here of dosi-
metric quantities for CT are of a provisional
nature.
Methods for assessing the patient dose of x rays
used in medical imaging are presented in Section 4.
Such methods are required for the determination of
normalization quantities used in dose-conversion
coefficients and of quantities recommended for
establishment and use of DRLs. Measurements of
these quantities rely mainly on the use of ionization
chambers or solid-state dosimeters, including TLDs.
For the low- and medium-energy x rays used in
medical imaging, the air kerma free-in-air is the
desired quality for calibration. Examples are given
of x-ray beamqualities recommended for calibration,
but it is noted that an international code of practice
for patient dosimetry in diagnostic radiology is pres-
ently being developed by the International Atomic
Energy Agency, including practical details of
calibrations.
Methods for determining organ and tissue doses
are the subject of Section 5. It is concluded that
assessment of organ and tissue doses, as well as
doses to the most heavily irradiated regions of the
body, mainly relies on Monte Carlo calculations.
Therefore, specific information is provided on the
application of Monte Carlo calculations of radiation
transport as employed for patient dosimetry in med-
ical x-ray imaging. This section includes comparis-
ons of dose-conversion coefficients calculated at
different institutes as well as comparisons of dose
measurements and calculations. It is noted that pro-
cedures for medical x-ray imaging vary from country
to country. Consequently, dose-conversion coeffi-
cients calculated by various authors commonly
refer to national or regional imaging procedures.
For similar exposure conditions, similar results are
obtained from calculations at different institutes
and from measurements. When a dose-conversion
coefficient is needed for a specific situation, the
best approach is to select a value from the available
International Commission on Radiation Units and Measurements 2005
data based on similarities in exposure conditions
(projection, view, and radiation quality) and a
patient model.
Appendix A presents information on backscatter
factors for irradiation conditions relevant for dosi-
metry for medical imaging using x rays. Appendices
B, C, and D provide dose-conversion coefficients that
reflect the differences in medical imaging using x
rays in the USA, Germany, and the UK, respect-
ively. Appendix E treats dose-conversion coefficients
for mammography. Appendix F describes a PC-
based Monte Carlo program for calculating patient
doses in radiography (excluding mammography) and
in fluoroscopy.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
8
Journal of the ICRU Vol 5 No 2 (2005) Report 74 doi:10.1093/jicru/ndi023
Oxford University Press
1 INTRODUCTION
1.1 EVOLUTION OF RADIATION DOSIMETRY
IN MEDICAL X-RAY IMAGING
The need for accurate measurement of x rays was
immediately evident in the early days of x-ray use in
medicine. This was not only because of the observed
biological effects induced by the exposure to x rays but
was also because of the instability in the generation of
x rays. The first measurement techniques made use of
the blackening of photographic emulsions and colour
changes of chemical compounds. However, no unit of
exposure existed and the development of instru-
mentation was, therefore, not very straightforward.
In 1905 at the Rontgen Congress in Berlin, a com-
mittee was founded to define a unit for the measure-
ment of Roentgen rays. Only after the suggestion by
Villard (1908) to make use of the change in electrical
conductivity of air by irradiation with x rays for their
quantification was a promising direction of experi-
mental attempts created. This resulted in 1925 in
the adoption, by the German Rontgen Society, of the
rontgen, R, as the unit of x-ray exposure. The same
year, at the First International Congress of Radi-
ology in London, a joint meeting of the Physics and
Radiology Sections was devoted to a discussion on
international standards for x-ray work. It resulted in
the appointment of the International X-ray Unit
Committee, with the mission to establish a uniform
x-ray standard of intensity and an x-ray unit. This
Committee became later on the International Com-
mission on Radiation Units and Measurements
(ICRU) (Taylor, 1958). The Second International
Congress of Radiology in Stockholm (1928) recom-
mended the definition of the Rontgen, which is in
essence the unit proposed by the German physicists,
in which the term Rontgen-ray dose was replaced by
the term quantity of radiation. The initial definition
of the unit was rephrased by the ICRU in 1931 and
1937: the essence of the unit was not changed but the
physical quantity of which the rontgen is the unit
had not been clearly defined (Allisy, 1995).
Early measurements with ionization chambers in
diagnostic radiology were difficult for two reasons:
(i) The first problem was the lack of sensitivity and
the strong energy dependence of the instruments.
(ii) The second was the lack of standards and ref-
erences to assist the interpretation of the res-
ults.
For further developments of x-ray dosimetry the
elaboration of an adequate standard dosimeter was
of major importance (Ku stner, 1924) and led to the
development of sophisticated and precise instru-
mentation based on ionization of air. The way to
dose quantification and understanding of biological
and technical aspects of x-ray diagnosis was opened.
Patient and staff dosimetry reflected the impact of
theories on biological effects at the time, resulting in
the use of derived quantities. Skin dosimetry with
the entrance skin exposure, related to cosmetic and
deterministic damage, was used to discuss limits of
patient exposure (Braun et al., 1928). For radiation
dosimetry, in the time period between the late 1920s
and the early 1950s the quantity exposure was
replaced by the quantity of absorbed dose (Taylor,
1990).
In the 1950s interest was focused on the induction
of genetic effects and cancer, mostly leukaemia. Con-
sequently, the anatomical sites included most often
in patient dose studies were the skin, the gonads,
and the active bone marrow (BM). The doses at the
skin could be directly measured with ionization
chamber dosimeters. The doses to the gonads and
the active BM were taken to be representative of
the totality of likely radiation effects. The absorbed
dose to the male gonads was usually taken to be the
same as that measured directly to the closely adja-
cent skin, whereas mean absorbed doses to the
female gonads or the active BM were related to
skin dose using the results of dose measurements
in physical phantoms. A large number of dosimeters
were required to obtain even approximate estimates
of the mean dose to widely distributed tissues such
as the active BM and many different exposure con-
ditions were needed to simulate the most common
types of x-ray examination. Nationwide surveys
using these dosimetric techniques combined with
studies of the numbers of x-ray examinations took
place in several countries in the period from 1955 up
to the early 1970s. Attention was focused on provid-
ing estimates of the genetically significant dose
International Commission on Radiation Units and Measurements 2005
(GSD) (UNSCEAR, 1972; NAS/NRC, 1980) and the
mean active BM dose to the population.
Developments in solid-state dosimetry led to the
increasing use of small unobtrusive thermolumines-
cent dosimeters (TLDs) directly attached to the
patient, for the measurement of the entrance surface
dose (ESD), including backscattered radiation, at
the centre of the x-ray beam. Alternatively, exposure
was measured with ionization chambers positioned
free-in-air on the x-ray beam axis. The results were
converted to the exposure at the patient entrance
surface plane using the inverse-square law. The res-
ults of measurements of radiation incident on the
skin were variously reported as entrance exposure
in rontgens (free-in-air) or as entrance absorbed dose
in rads (ICRU, 1961). The material in which the
absorbed dose was measured was air, water, or soft
tissue. Measurements were made either free-in-air
or on the surface of a patient or a phantom. There is
approximate numerical agreement (to within 15 %)
between exposures expressed in rontgens and
absorbed dose to air, water, or soft tissue expressed
in rads, for the x-ray spectra used in medical
imaging. However, doses measured with the patient
or phantom present can be up to 60 % higher than
those measured free-in-air at the skin-entrance
plane because of the contribution of backscattered
radiation.
In the 1960s a special type of large-area transmis-
sion ionization chamber dosimeter became available
to measure the radiation incident on the patient in
terms of the product of the exposure and the area of
the x-ray beam in units of R cm
2
(Arnal and Pychlau,
1961; Morgan, 1961; Cameron, 1971). The chamber
intercepted the entire useful beam, irrespective of
the collimator setting, so that its response was pro-
portional to both the area of the beam and the expos-
ure. This was thought to provide a more nearly
complete measure of the total exposure of the patient
and hence to be more closely related to the radiation
risk, which depends on the extent of the irradiated
volume within the patient as well as the exposure at
the centre of the x-ray beam. Moreover, there is
considerable practical advantage in the flexibility
afforded in the positioning of the chamber owing to
the approximate invariance of the product of expos-
ure and beam area for all planes perpendicular to
the beam axis between the beam defining collimat-
ors and the patient. However, this device led to the
introduction of a quantity and unit (exposurearea
product and R cm
2
), unfamiliar to the practitioners
at that time, to add to those already used to express
the dose to the patient in diagnostic radiology.
It was soon appreciated that measurements of the
exposurearea product (or strictly the integral of the
exposure over the area of the beam), together with
data on the x-ray spectrum, could be readily conver-
ted into the total radiant energy in the beam. Hence
it could be used to make estimates of the total energy
imparted to the patient or the integral dose
(Carlsson, 1963) as it was then known. The integral
dose, defined as the mass integral of the absorbed
dose over the total body (expressed in units of g rad,
or erg), was seen to be more closely related to the
likely biological effects than the exposurearea
product. It had the considerable advantage over
organ doses that it could be derived directly from
exposurearea product measurements. As long as
other details of the exposure conditions were avail-
able, it allowed a reliable estimate of the energy that
was deposited in the patient. This fraction is critic-
ally dependent on the x-ray beam quality, the size
and position of the x-ray beam in the patient, the
thickness of the part being x-rayed, and the propor-
tion of the beam intercepted by the patient. Because
the integral dose was often easier and more practical
to estimate than the doses to individual organs,
many studies in the 1960s and 1970s reported
doses to patients in this way.
In the 1980s the names and definitions of the
quantities and units used in diagnostic radiology,
as in other areas of radiation dosimetry, underwent
major changes following the publication of revised
recommendations by the ICRU (1980), which advoc-
ated adoption of the International System of Units.
As a result, exposure was replaced by air kerma (AK)
(unit: joule per kilogram with the special name
gray) as the quantity in which dosimeters were cal-
ibrated and linked to the national primary stand-
ards. Absorbed doses were also expressed in gray,
thus improving the numerical agreement between
the quantity measured by dosimeters and the
absorbed dose to soft tissue (to within 5 %for medical
x rays and most tissue-equivalent materials).
Exposurearea product (R cm
2
) was replaced by
kermaarea product (Gy cm
2
) or (absorbed)
dosearea product (DAP) (also Gy cm
2
) and integral
doses were expressed as the total energy imparted to
the patient in joules.
The limitations of the concept of the total energy
imparted to the body and its relation to possible
radiation effects became more apparent when com-
putational methods of dosimetry became more
advanced. They allowed the mean absorbed doses
to individual tissues and organs of relevance to radi-
ation protection to be calculated. Using simple geo-
metrical models of the body and of typical medical
x-ray beams, Monte Carlo radiation transport codes
enabled researchers to calculate organ doses nor-
malized to easily measured quantities such as the
entrance skin exposure, the ESD, or the DAP for a
wide range of medical x-ray exposure conditions.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
10
Better information on the relative radiosensitivities
of different organs and tissues became available in
the late 1980s (UNSCEAR, 1988; BEIR, 1990; ICRP,
1991b). This information combined with knowledge
of the mean absorbed doses to these organs and
tissues led to estimates of the total radiation risk
from the partial body exposures typical of medical
imaging.
There is, however, a need for a single dosimetric
quantity related to the total potential health detri-
ment to provide a practical tool for optimization of
protection for the patient. Unless one particular
organ is completely dominant in determining the
totality of likely health effects, it is inconvenient to
assess the risk from a list of organ doses. The dose to
some organs may be increased and to others
decreased by the technique changes under consid-
eration. For practical comparisons of radiation risks
for different techniques or procedures, some invest-
igators have applied the concept of effective dose a
weighted sum of organ doses developed by the ICRP
primarily for use in its system of radiological protec-
tion(ICRP, 1991b) to medical exposures. However,
effective dose should not be used, for example by
using the nominal fatality coefficients (ICRP, 1991b),
for assessment of detriment from exposure due to
medical x-ray imaging. Such assessments could be
inappropriate because of potential differences in
health status, gender, and age between a particular
group of patients and the reference population for
whom the ICRP derived the risk coefficients. Fur-
thermore, the conditions of a low dose and especially
a low dose rate, assumed by the ICRP, may arguably
not always be met in practice in radiology.
1.2 RISKS FOR THE PATIENT IN
RADIOLOGICAL IMAGING AND RELEVANT
DOSIMETRIC QUANTITIES
The objective of dosimetry in radiological imaging
is the quantification of radiation exposure within an
approach to optimize the image quality to absorbed
dose ratio. The image quality should be understood
as the relevant information appropriate to the
clinical situation. Dosimetry also provides the
means to avoid excessive doses that could imply a
significant risk of induction of deterministic effects,
for example, for some cases in interventional radi-
ology. The dosimetric quantities and dosimetric pro-
tocols relevant in radiological imaging are those
most closely related to the major (or more frequent)
risks for the patient.
Most of our knowledge on induced radiobiological
effects is based on the relation between absorbed
dose and biological effect. For radiation protection
and therapy applications, the absorbed dose has
sometimes to be weighted by appropriate factors to
predict the effects or risks (Hall, 2000; Wambersie
et al., 2002; Zoetelief et al., 2003a). International
recommendations and national and international
regulations are also based on absorbed dose or
weighted absorbed dose (ICRU Report 60, 1998c;
ICRP Publication 60, 1991b; IAEA, 1996a; EC,
1997).
Figure 1.1. Acute dermatitis 3 weeks after excessive skin
irradiation (several Gy) during an interventional procedure
(placement of a porto-cave shunt). The lesion is well delineated
(12 cm in diameter) and located at the back of the patient, in the
upper right paralumbar area. One can notice the beginning of
re-epithelialization originating from the border of the lesion
(arrows 1) but also within the irradiated area. Inside the irradi-
ated area, each white spot is a clone resulting from the prolifera-
tion of a surviving basal stem cell of the epidermis (only two
clones are shown by arrows 2) [P. Goffette, Imaging
Department, UCL University Clinics St Luc, Brussels, cited in
Wambersie et al. (2005), by permission of Oxford University
Press].
Figure 1.2. Chronic dermatitis on the skin of the same patient as
in Figure 1.1. One year after irradiation, sclerosis of the dermis,
numerous telangiectasiae, total loss of elasticity, fixation to the
deep tissue layers, frequent ulcerations [P. Goffette, Imaging
Department, UCL University Clinics St Luc, Brussels, cited in
Wambersie et al. (2005), by permission of Oxford University
Press].
INTRODUCTION
11
1.2.1 Acute deterministic effects
1
In the early years of application of x rays in radi-
ology, several cases of acute and severe skin reac-
tions were reported. With the improvement of the
equipment, the use of safer working procedures, and
better training, the number and severity of these
accidents decreased rapidly and even became excep-
tional. In recent years, however, with the rapid
development of interventional radiology, acute skin
reactions became again a major concern for patient
safety (Figures 1.1 and 1.2).
At high dose (>2 Gy), the severity of the lesions
can be predicted from the local absorbed dose in the
skin and underlying dermis. Therefore, the skin
dose in the most heavily irradiated area is usually
the most relevant quantity to be determined in
interventional radiology. This implies that both the
maximum dose as well as the surface irradiated
above the tolerance dose be evaluated (ICRP Pub-
lication 59, 1991b). In interventional radiological
procedures, one of the main practical issues is often
the identification of the skin region that receives the
highest dose, as discussed in Section 3.
1.2.2 Late effects
1.2.2.1 Cancer induction
Cancer induction is generally considered to be the
main risk for patients after radiological imaging.
The quantitative approach of the risk assessment is
a more complex issue than for acute effects, because
we are dealing with (rather low) probabilities at low
doses (UNSCEAR, 2000). The risk of cancer induc-
tion in an organ or tissue is assumed to be related to
the average absorbed dose in that organ and
strongly depends on the type of organ or tissue. In
radiation protection, the relative risk of lethal can-
cer induction is assumed to be the product of the
organ dose and a weighting factor (w
T
), which
expresses the particular susceptibility of that organ
to cancer induction (Section 3.3; ICRP Publication
60, 1991b).
For the purpose of radiation protection, i.e., for the
management of radiation risk, the total risk of lethal
cancer induction for a person is assumed to be
related to the sum of the weighted organ doses.
This is known as the concept of effective dose, mainly
designed for occupational exposure (ICRP Pub-
lication 60, 1991b). The numerical values of the
weighting factors were selected for an average
adult population. In radiology, more appropriate
weighting factors could probably be selected taking
into account the characteristics of individual patients
undergoing radiology such as gender and age (e.g.,
the variation of susceptibility of the breast for breast
cancer induction).
Cancer induction is generally considered to be a
stochastic effect with the probability (and not the
severity) of effect increasing with dose. For cancer
induction, however, this traditional distinction
between deterministic and stochastic effects is likely
to have limitations. There is indeed evidence that
chronic radiation-induced lesions, such as sclerosis
or chronic dermatitis, are particularly dangerous
pre-cancerous lesions, although they are clearly
deterministic effects (Boice et al., 1985, 1988; Arai
et al., 1991; ICRP Publication 59, 1991a; Wambersie
et al., 1995).
In addition to the lethal cancers taken into
account in the weighting factors discussed above,
the incidence of non-lethal cancers is of importance.
1.2.2.2 Late effects in normal tissues
An increasing number of data indicates that some
late effects in normal tissues, such as the heart,
represent a risk equally important as the risk of
cancer induction, for doses larger than several hun-
dred mSv (Brenner et al., 2003; Preston et al., 2003;
Tubiana et al., 2005).
1.2.2.3 Impairment of mental development
Evidence has been reported that doses such as
those delivered by repeated computed tomography
(CT) examinations performed in sub-optimal condi-
tions in children may significantly impair mental
development (Hall, 2002; Hall et al., 2004; Yamada
et al., 2004).
1.2.2.4 Genetic risk
For some radiological procedures involving the
gonads, the genetic risk has to be considered, espe-
cially for younger patients. A careful analysis, based
on a critical review of the available data, however,
concluded that the previous evaluation of the risk
was significantly overestimated. It has now become
possible to estimate risks for all classes of genetic
diseases (which was not the case until 1993) and that
the risks are small compared with the baseline
risks (Sankaranarayanan and Chakraborty, 2000;
UNSCEAR, 2001).
1.2.3 Relevant dosimetric quantities and
dosimetric procedures
On the basis of the discussion of the main
radiation-induced effects and risks, the following
1
The ICRP has recently proposed (2005) to replace the
term deterministic effects by tissue reactions.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
12
relevant dosimetric quantities and dosimetric pro-
cedures can be derived.
Absorbed dose, expressed in gray (Gy), is the rel-
evant quantity to be determined and reported in
radiological imaging. In practice, it should be
determined in water (or water equivalent) and simu-
lating or using patient geometry, i.e., taking into
account scattered radiation and shape of patient
contour.
Absorbed dose should be evaluated at the level of:
Skin, because the skin is the most heavily
exposed tissue. Reports on skin dose should
include the site location, maximum dose, and
skin surface irradiated above the tolerance level
that should be specified;
The most heavily irradiated organs, taking into
account the specific susceptibility of the organs for
cancer induction and late effects as discussed
above.
Depending on the radiological procedure that is
performed, dose should be evaluated for the follow-
ing organs: female breast, heart, thyroid, gonads
(depending on age), brain (especially in children),
and also dose to the embryo. The maximum dose
and the average organ dose should be evaluated.
The susceptibility of the female breast for cancer
induction strongly varies with age. The average
dose to the glandular tissue is probably more signi-
ficant than the average dose to the organ as a whole.
The use of CT raises specific problems that will be
discussed in Section 3.2.
1.2.4 Required accuracy
The required accuracy for dose determination
depends on the dose level and potential risk. An
accuracy requirement of 7 %, at an expanded uncer-
tainty using a coverage factor of 2 corresponding to
the confidence level of 95 %, is recommended for
comparative risk assessments as well as for quality
assurance (AAPM, 1992). When there is a risk of
deterministic effects similar accuracy will be needed.
For calibration the expanded standard uncertainty
should not exceed 5 % (AAPM, 1992). These require-
ments can, in general, be achieved.
An accuracy of 3050 % can be accepted in the
cases where organ doses are low. For comparison,
in occupational exposure or radiation protection, an
accuracy of 50 % may also often be accepted in some
situations. In contrast, in radiation therapy, an
accuracy on dose, at the reference point, better than
5 % is aimed for in the case of radical treatment.
In the case of an accidental acute effect (e.g., skin
burn), sufficient information should be available to
the staff of a department of radiology to allow dose
reconstruction with a reasonable accuracy to predict
the severity of the possible harm.
1.3 DOSIMETRY IN RADIOLOGY: RELEVANT
QUANTITIES
1.3.1 Calibration at the Standards
Laboratory
In diagnostic radiology, dosimetry is based on
calibrated instrument(s), usually ionization cham-
ber(s). These instruments (used as local reference
in the department) are calibrated at National
Standards Laboratories, or at laboratories with
instruments directly traceable to National Stand-
ards Laboratories. The calibration coefficients in the
certificates are expressed in terms of air kerma free-
in-air. The manufacturers also express their calibra-
tions in these terms. Any change in the situation is
unlikely in the foreseeable future.
1.3.2 From air kerma free-in-air to
absorbed dose in water in patient or
phantom
Conversion coefficients are used to obtain
absorbed doses in organs or tissues and at selected
clinically relevant points (Figure 1.3). These conver-
sion coefficients depend on a number of factors that
are discussed in detail in the present report. They
are based in part on ICRU Report 57 (ICRU, 1998a).
1.3.3 Air kerma-area product (KAP) and
dosearea product (DAP)
The output of an x-ray tube can be monitored with
a transmission ionization chamber. The chamber is
mounted on radiological tubes, downstream relative
to the collimator (Figure 1.3). The increasing use of
this monitoring device is partly a consequence of the
EC Directive (EC, 1997).
The signal of the chamber is calibrated in terms of
KAP (or absorbed DAP), at a specified distance free-
in-air. It is expressed in mGy m
2
. As more convenient
multiples, the radiology community would prefer
Gy cm
2
. The transmission chamber acts as a
monitor, like in radiation therapy, and should be
calibrated against the reference dosimeter of the
department.
Fromthe radiological parameters (Section1.3.4.1),
the readings of the reference chamber, and the KAP
(or DAP), it is possible to derive useful information:
Doses at reference points, for example, entrance
(skin) dose on the beam axis at a given distance,
and at specified points in organs.
Mean organ doses.
INTRODUCTION
13
This, however, requires sophisticated Monte Carlo
programs, which produce conversion coefficients
valid only for well-defined procedures and within
strict (geometric) limits that should be specified
(Chapter 5; Appendices B, C, D, and F; Struelens,
thesis, 2005).
As part of a quality assurance program, it is
important to check regularly the reproducibility of
the response of the transmission chamber with a
calibrated dosimeter.
A normalization quantity is a dosimetric quantity
that can be readily measured or calculated in the
clinical situation. Normalization quantities are used
to derive aspecifieddosimetric quantity, for example,
mean organ dose, using an appropriate conversion
coefficient. To date KAPor DAP, (Section 2), are com-
monly used as normalization quantities, but various
other quantities can also be used (Section 3.4).
1.3.4 Reporting patient irradiation in
radiological imaging
1.3.4.1 Radiological parameters of the exposure
Radiological parameters, such as tube voltage
(kV), tube currentexposure time product (mAs),
filtration, exposure time (for fluoroscopy), field size
and position, and irradiated region, should be recor-
ded for the standard medical imaging procedures. In
recent equipment, in particular in direct digital radi-
ology systems (flat panel detectors), recording of
these parameters is made automatically (Figure 1.4).
To the extent that the tube output remains con-
stant relative to these parameters, this set of
information allows evaluating doses at reference
points where previous calibration has been per-
formed at the surface or inside a phantom. Informa-
tion obtained on the phantom can then be used to
evaluate the dose to the patients, but only for routine
procedures and with the necessary care.
The ICRU recommends that the radiographic con-
ditions of the irradiations be reported as completely
as possible. When these are carefully reported, it will
be possible later on to apply better conversion coef-
ficients when available and to derive more relevant
quantities.
1.3.4.2 Air kermaarea product (KAP) or
dosearea product (DAP)
KAP or DAP provides a continuous monitoring of
the x-ray tube output and an indication of the
absorbed dose at reference points at the skin and
the possibility to calculate organ doses for standard
Figure 1.3. Calibration of dosimeters to be used in a diagnostic radiology. (1) The reference dosimeter used in diagnostic radiology is
calibrated (directly or through transfer dosimeters) at a Standards Laboratory in air kerma free-in-air (left). (2) In diagnostic radiology,
the dosimeter can be used to measure the entrance surface dose (skin dose) provided appropriate conversion coefficients are applied, i.e.,
from air kerma free-in-air to absorbed dose in water including backscatter (right). In addition to skin dose, the chamber can also be used to
determine the dose at any point in the patient. (3) The transmission chamber, regularly calibrated (Section 4.2.4), is used as a monitor. It
provides the DAP or the air kermaarea product (KAP), depending on how it is calibrated. From the radiological parameters (Section
1.3.4.1), the readings of the reference and KAP dosimeters, in principle, organ doses can be computed for any organ using complex Monte
Carlo programs. The resulting conversion coefficients are valid within strict limits and specific for each radiological procedure (from
Wambersie et al., 2006, by permission of Oxford University Press).
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
14
procedures. KAP or DAP can be used as normaliza-
tion quantities to derive relevant dosimetric
quantities, such as organ doses, using conversion
coefficients (Sections 1.3.3 and 3.4).
Continuous monitoring of KAP relative to the
radiographic parameters of exposure listed previ-
ously (Section 1.3.4.1) improves the confidence in
both approaches.
1.3.4.3 Monte Carlo computation
Sophisticated Monte Carlo programs have been
designed to calculate doses at specified points or
average organ doses in phantoms starting from the
radiographic parameters and/or the KAP or DAP
(Figure 1.5). The results from these approaches are
available mainly for standard, well-established
radiological procedures. The use of such Monte
Carlo codes needs considerable expertise when
applied for specific or uncommon procedures.
1.3.4.4 Phantoms and in vivo measurements
Different types of phantoms have been designed
(ICRU Report 44, 1989b; ICRU Report 48, 1992b).
Transfer of doses derived from measurements using
phantoms to doses for individual patients or patient
series requires appropriate care, but some programs
allow the introduction of individual patient para-
meters (gender, age, shape, dimensions, organ size
and location, etc.) (Tapiovaara et al., 1997).
Patient measurements may be considered as a
validation (in clinical conditions) of phantom meas-
urements to derive the doses. However, in radiology
such measurements are often not a realistic scenario
for every type of procedure on a systematic base.
They are, however, strongly recommended on a lim-
ited series of patients for standard procedures.
1.3.5 Discussion
In radiological imaging, one of the dosimetric
issues is the fact that several quantities are used to
quantify the magnitude of the exposure of the
patient to ionizing radiation. There is some ambigu-
ity (and even disagreement) in their names and
applications in radiological procedures (e.g., kerma
versus absorbed dose, air versus water, free-in-air or
with backscatter).
The numerical values of the different quantities
may be close to each other and within dosimetric
uncertainties (even within clinical accuracy
Figure 1.4. Correlation between radiological parameters, transmission monitor readings [air kermaarea product (KAP) or dose area
product (DAP), and clinically relevant dosimetric quantities (see arrows 3 and 1, respectively). This correlation should be checked at
regular intervals depending on the type of radiological examination, the irradiated organs and level of exposures, and the type of patients
(pregnant women, children, etc.]. The radiological parameters include tube voltage (kV), tube currentexposure time product (mAs),
filtration, field size (at a given distance), and exposure time for fluoroscopy (Section 1.3.4.1). The clinically relevant dosimetric quantities
include the skin (entrance) dose on the beam axis, skin surface (and regions) irradiated above the tolerance dose for induction of
deterministic effects at clinically relevant points and/or organs. Left part Regular checks of the correlation between the radiological
parameters and the readings of the KAP (or DAP) dosimeter (arrows 2) guarantee output stability and the reliability of both dosimetric
approaches. It improves the confidence in the two sets of dosimetric results. Right part In the absence of a KAP (or DAP) meter, regular
checks of the correlation (arrow 3) between the radiological parameters and the readings of the reference ionization chamber, at reference
points, may also provide some guarantee on the stability and reliability of the dosimetry. This latter procedure, however, should be limited
to simple, routine, and well-defined radiological examinations and is not recommended for examinations of pregnant women and children.
Courtesy: A. Wambersie.
INTRODUCTION
15
requirements). This cannot be used as an excuse for
not specifying clearly the involved quantities.
There is still debate concerning the relevant
quantity to select in radiology: (i) the standards
laboratories (and manufacturers) calibrate the ion-
ization chambers in air kerma free-in-air; (ii) the
radiological community is using the concept of
absorbed dose; (iii) a device informing the practi-
tioner of the quantity of radiation produced by the
equipment during the radiological procedure (in
practice a transmission chamber) is now imposed
by regulation in several countries (EC, 1997).
This situation is very similar to the situation in
radiation therapy. In each radiation therapy depart-
ment, there is a reference dosimeter (ionization
chamber) calibrated for
60
Co gamma rays or a few
MV x rays. Conversion coefficients are available for
photon and electron beams of higher energies (ICRU
Report 64, 2001; ICRU Report 71, 2004). A transmis-
sion chamber used as a monitor is regularly calib-
rated against the reference dosimeter in specified
conditions.
It is the responsibility of the local medical physics
team to select and apply the appropriate con-
version coefficients and to check the reproducibility
of the responses of the different dosimetric
systems. Similar working procedures and sharing
of responsibilities can be applied in a radiology
department.
The medical and radiobiological community, radi-
ologists, and radiation oncologists currently relate
the biological effects to the absorbed dose. A large
experience has been built using absorbed dose as a
quantification of the magnitude of exposure and
dose-effect relationships. It would not be safe or
wise to modify this well-established approach.
When using ionization chambers calibrated in AK,
it is thus important to select the appropriate conver-
sion coefficients. In that respect, the ICRUReport 74
where this issue is carefully discussed is thus timely.
In the European and national regulations and
recommendations, the quantities absorbed dose and
weighted absorbed dose (equivalent dose, effective
dose, etc.) are used. This may interfere with the
Figure 1.5. Visual representation of the photon tracks (primary and scattered radiation) and energy distribution (colour code: light gray,
2040 kV; medium gray, 4060 kV; dark gray, 6080 kV) of a Monte Carlo simulation for a PA view of a pelvic exposure, as part of an
angiographic examination of the lower limbs. The exposure was assumed to be performed at a peak tube voltage of 80 kV, a total filtration
of 4.6 mm Al
eq
, and a focus-to-skin distance of 55 cm. As model for the patient, the mathematical phantom BODYBUILDER was used.
Conversion coefficients can be retrieved between different calculated values, such as a specific organ dose or DAP/KAP. Reproduced with
permission from Dr. L. Struelens.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
16
selection of the quantities in the radiological depart-
ments (EC, 1997; ICRP Publication 73, 1996).
Dosimetry for CT raises specific issues. For CT,
skin dose and dose at a point do not have the same
significance as for conventional radiography, and
specific indexes thus need to be introduced. The CT
dose index (CTDI) has been defined for the assess-
ment of organ dose and QA (Sections 3.2 and 4.2.5).
Several methods were published for the practical
assessment of CTDI (IEC, 1999; EC 2000).
For interventional radiology, where the avoidance
of deterministic effects becomes important, dose-
conversion coefficients relating the absorbed
dose to the more heavily irradiated site at the
surface of the body to normalization quantities
(Section 3.4) are not yet commonly available, except
for some coronary procedures. In particular, the
point (place on skin) where the maximum dose is
obtained may not always be known in advance
(Section 3.3.1).
1.4 NEED FOR HARMONIZATION OF
QUANTITIES AND TERMINOLOGY
Various quantities and terminologies have been
used for the specification of dose on the central
beam axis at the point where the x-ray beam enters
the patient (or a phantom representing the patient).
These include the exposure at skin entrance (ESE),
the input radiation exposure, the entrance surface
air kerma (ESAK), the entrance air kerma, the AK,
the ESD, the entrance skin dose (ESD), and the
integral skin dose (ISH and EC, 1998).
ICRU Reports 33 and 60 (1980, 1998c) recommen-
ded the use of the International System of Units (SI).
In ICRU Report 33 it is stated that the special unit
of exposure, the rontgen, should be dropped by
1985 and be replaced by the SI unit C/kg. Con-
sequently, the approximate numerical equivalence
of exposure, AK, and absorbed dose to air was lost.
As a result of the introduction of the SI, exposure has
been replaced by air kerma (free-in-air) as the
quantity in which dosimeters are calibrated at
standards laboratories.
Sometimes it seemed to be the convention that
kerma implies absence of backscatter and absorbed
dose implies the presence of backscatter. However,
the presence or absence of backscatter cannot be
derived from the definitions of kerma and absorbed
dose (ICRU, 1998c). Calibrations of dosimeters are
generally made in terms of air kerma free-in-air.
Therefore, it is often assumed that kerma is
expressed only in air. However, the ICRU (1998c)
explicitly states that one can refer to a value of
kerma for a specified material at a point in free
space, or inside a different material. Different
names are used in practice for the same quantity,
for example, entrance surface air kerma, air kerma,
and entrance air kerma. The same abbreviation ESD
is used for both entrance surface dose (absorbed dose
determined free-in-air most likely expressed in air)
and entrance skin dose (absorbed dose most likely
expressed in skin tissue).
The kermaarea product is often used for dose
assessment for more complex examinations, includ-
ing radiography and fluoroscopy. Although the name
does not state this explicitly, the kerma is usually
expressed in air free-in-air and the backscatter from
a patient or a phantom is not to be included.
In interventional radiology, the value of the
kermaarea product for a complete examination
has been used as an indicator for the occurrence of
stochastic effects, whereas information on the max-
imum dose at skin entrance (that is, the dose at the
location of the skin where it is highest) is of import-
ance with respect to the possible occurrence of deter-
ministic effects. The maximum dose at skin entrance
is also referred to as the peak skin dose (Miller et al.,
2002) and defined as the highest dose delivered to
any portion of the patients skin.
For the assessment of organ doses and quality
assurance in CT, the CTDI has been defined as the
integral of the absorbed dose profile along a line
parallel to the axis of rotation of the scanner divided
by the nominal slice thickness (Shope et al., 1981). In
the literature, different methods can be found for the
practical assessment of CTDI. These include differ-
ences in the boundaries of integration, the use of a
phantom or measurement free-in-air, and differ-
ences in the material in which the absorbed dose is
expressed, for example, polymethylmethacrylate
(PMMA) or air.
The present situation in patient dosimetry for
medical x-ray imaging clearly indicates the need
for dose quantities recommended for the different
applications and the need for using the same, self-
consistent, system for names, symbols, and units.
1.5 THE TWO PURPOSES OF
PATIENT DOSIMETRY
In medical x-ray imaging there are two funda-
mental reasons for measuring or estimating the
patient dose. First, measurements provide a means
for setting and checking standards of good practice,
as an aid to the optimization of the radiation protec-
tion of the patient and of image quality. Second,
estimates of the absorbed dose to tissues and organs
in the patient are needed to assess radiation
detriment so that radiological techniques can be
justified and cases of accidental overexposure
investigated.
INTRODUCTION
17
1.5.1 To set and check standards of
good practice
It is well known from the results of numerous
surveys that there is considerable variation in the
doses delivered to patients from the same type of x-
ray examination conducted in different facilities or
even within a single facility. Without some form of
patient dose monitoring, it is difficult to know the
performance of an x-ray facility and to judge how it
compares with generally accepted practice. As an aid
to the optimization of the radiation protection of the
patient, reference values, variously called reference
dose values (EC, 1996c), DRLs (ICRP, 1996), or guid-
ance levels (IAEA, 1996a), can be specified for par-
ticular x-ray imaging tasks. Local performance can
be checked against these reference values by peri-
odic measurement as part of a quality assurance
program. For these purposes, clearly defined quant-
ities are required, which can be easily measured
with readily available instruments of sufficient pre-
cision and accuracy. Consequently, dosimetric
quantities associated with the primary x-ray beam
(e.g., AK at a specified point on the beam axis) or
closely related radiation quantities suited to particu-
lar applications (e.g., KAP, CT kerma index or CTDI)
have become established quantities used to set and
check standards of good practice.
1.5.2 To assist in assessing detriment or
harm
For the justification of a practice or for the invest-
igation of cases of accidental over-exposure, the
absorbed dose to the patient should be expressed in
a manner that can be directly related to the potential
health risks. For stochastic radiation effects, the
ICRP (1991b) has recommended that the mean
absorbed dose to individual organs or tissues is the
quantity of interest. If deterministic effects are con-
sidered possible, for example, in some cases of inter-
ventional radiology, the absorbed dose to the more
heavily irradiated sites at the surface of the body,
such as the skin in the primary beam, is the radi-
ation quantity of interest.
Organ or tissue doses cannot be directly measured
in patients but can be derived from other directly
measurable radiation quantities using appropriate
conversion coefficients. Extensive tabulations of
such conversion coefficients for reference patients
and reference irradiation conditions have been pub-
lished, are available in computer readable form, or
can be calculated (Appendices BF). These coeffi-
cients relate organ doses to practically measurable
normalization quantities, some of which are the
same as the quantities used for setting and checking
standards of good practice. Alternatively, organ and
tissue doses can be derived from measurements
inside physical phantoms.
1.6 RELATIONSHIP BETWEEN PATIENT
DOSE AND IMAGE QUALITY
It is important to ensure that efforts to reduce
patient doses do not also reduce doses to the image
receptor to such an extent that the quality of the
images is degraded to an unacceptable level. Image
quality can be affected by inadequate doses in four
distinct ways:
(i) In the non-digital imaging systems used in con-
ventional radiography and fluoroscopy, optical
density or brightness of the image depends on
the dose and dose rate received by the image
receptor, respectively. Too low a dose or dose
rate can result in images that are too faint to be
clearly discerned.
(ii) Dose reduction by increasing the tube voltage
and thereby allowing a reduction in tube cur-
rent or exposure time to maintain the same
dose to the image receptor can degrade image
quality by decreasing contrast.
(iii) As medical imaging systems have become more
sensitive, needing only lower doses to achieve
images of satisfactory density or brightness,
there is an increased likelihood that random
variations in the photon fluence rate reaching
the image receptor will give a disturbing
mottled appearance to the image. This so-
called quantum mottle is preferably the dom-
inant source of image degradation in sensitive
digital and non-digital imaging systems.
(iv) The sensitivity of the imaging system can often
be improved by increasing the thickness of the
sensitive layer of the image receptor so that it
absorbs more of the incident x-ray energy. For
the majority of image receptors that re-emit the
absorbed energy in the form of visible light,
thicker sensitive layers result in wider spatial
dispersion of the emitted light before the image
is recorded. Greater sensitivity, and hence the
possibility of using lower doses, is consequently
gained at the expense of poorer spatial resolu-
tion in the image.
Procedures for checking that doses have not been
reduced to such an extent that inadequate optical
density, excessive noise, poor spatial resolution, or
lack of contrast prevents reliable diagnosis should be
an essential component of x-ray department quality
assurance programs. A range of phantoms that can
be used to assess image quality in diagnostic radi-
ology are described in ICRU Report 48 (1992c). More
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
18
fundamental methods for assessing the quality of
medical images are published by Doi et al. (1982;
1986) and presented in ICRU Reports 41 and
54 (1986, 1995). In addition, an ICRU report has
been recently published on Chest radiography-
Assessment of Image Quality (ICRU Report 70,
2003) and one is underway in Mammography
Assessment of Image Quality.
1.7 SCOPE OF THE REPORT
The present Report is primarily concerned with
methods of patient dosimetry in medical x-ray
imaging. Methods are described for dosimetry per-
formed to set or check standards of good practice and
as a step towards the assessment of detriment or
harm. In many situations the dosimetric quantity
of interest can be directly measured. In other cases,
however, it is not practicable to measure a quantity
of interest directly (e.g., organ or tissue dose).
Instead it is obtained indirectly by application of an
appropriate conversion coefficient to a quantity that
can be measured directly.
All aspects of patient dosimetry require know-
ledge of the properties of the x-ray beam. The basic
features of the x-ray beams used for medical imaging
and their specification and measurement are, there-
fore, discussed in Section 2. The various dosimetric
quantities to be used for patient dosimetry are
defined and discussed in Section 3, and appropriate
notations are introduced. The discussion is divided
into five parts: basic dosimetric quantities,
application specific quantities, risk related quantit-
ies, dose-conversion coefficients, and quantities
recommended for establishment and use of DRLs.
Section 4 discusses the methodology to be used for
the measurement of specific dose quantities includ-
ing incident AK, ESAK, KAP and various quantities
to be used for CT dosimetry. It considers the choice
and calibration of dosimeters, the practical measure-
ment technique, the measurement uncertainty, and
the advantages and disadvantages of phantom-
based and patient measurements. The derivation of
backscatter factors, which can be used to relate
incident AK and ESAK to ESD, is treated in
Appendix A. The conversion coefficients used to
estimate organ and tissue doses are covered in Sec-
tion 5. Two approaches to determine the conversion
coefficients are considered: measurements in phys-
ical phantoms and computational methods. The lat-
ter use Monte Carlo techniques to simulate radiation
transport through computer-based models of the
patient. This is usually the more useful approach.
The main features of the computational method and
the associated uncertainties in the conversion coef-
ficients are discussed. Results of such calculations
are available from a number of sources for conven-
tional projection imaging, CT, and mammography.
Conclusions, including recommendations on the
selection of the most appropriate conversion coeffi-
cient for a particular examination or procedure, are
presented in Section 6. Organ dose-conversion coef-
ficients from different sources are considered in
Appendices BF.
INTRODUCTION
19
Journal of the ICRU Vol 5 No 2 (2005) Report 74 doi:10.1093/jicru/ndi024
Oxford University Press
2 SPECIFICATION OF X-RAY BEAMS
X-ray tubes used in medical x-ray imaging employ
peak voltages of between 25 and 150 kV to acceler-
ate electrons from a heated cathode towards an
anode target. X rays are produced either as a result
of radiative energy loss by electrons slowing down in
the target material (bremsstrahlung) or by radiative
transitions of electrons within the atoms of the
target (characteristic x rays). The former have a
continuous range of energies from that of the most
energetic electron downwards whereas the latter
occur at discrete energies characteristic of the target
material.
The radiation quality of an x-ray beam can be
characterized by the x-ray spectrum. X-ray spectra
can be measured by using spectrometers based on
scintillation counters, germanium or silicon detect-
ors, or by crystal diffraction. These techniques,
however, require considerable expertise and are
time-consuming to perform. Therefore, it is recom-
mended that the radiation quality of x-ray beams
used for medical imaging be characterized by a com-
bination of various parameters. These include first
half-value layer (HVL; symbol HVL
1
); the second
HVL (HVL
2
); the ratio of HVL
1
to HVL
2
, referred to
as the homogeneity coefficient; the tube voltage and
the total filtration. In most cases, the quality of an
x-ray beam can be adequately specified by means of
the combined information on tube voltage, HVL
1
,
and HVL
2
, or the tube voltage, HVL
1
, and total
filtration.
The yield is also an important characteristic of an
x-ray tube (including filtration). For this purpose the
x-ray tube output is defined in Section 2.5.
2.1 PHOTON SPECTRUM
The shape of the spectrum depends on the voltage
applied to the x-ray tube (kV), the waveform of
the generator, the target material and angle, and
the amount of inherent and added filtration in the
x-ray beam. For most diagnostic examinations, and
all interventional procedures, x-ray spectra from a
tungsten target are used (Figure 2.1). Aluminium
filtration is generally employed to remove the
low-energy end of the bremsstrahlung spectrum,
which would otherwise be absorbed in the superficial
tissues of the patient without contributing to the
final image. An additional filter of, for example, cop-
per may also be used in some situations to further
harden the spectrum. To prevent the low-energy
characteristic x rays produced in the copper filter
from reaching the patient the aluminium filter
should be between the copper filter and the patient.
There are peaks in the tungsten spectra due to K
a
and K
b
characteristic radiation for peak tube vol-
tages >69.5 kV, corresponding to the energy of the
0 20 40 60 80
PHOTON ENERGY; E/keV
R
E
L
A
T
I
V
E
N
U
M
B
E
R
O
F
P
H
O
T
O
N
S
0 20 40 60 80
PHOTON ENERGY; E/keV
R
E
L
A
T
I
V
E
N
U
M
B
E
R
O
F
P
H
O
T
O
N
S
Figure 2.1. Typical photon fluence x-ray spectra used for medical
imaging (IPEM, 1997). Top tube voltage: 80 kV constant poten-
tial, tungsten anode; filtration: 3 mm Al, anode angle 16

; HVL
1
:
2.98 mm Al. Bottom tube voltage: 28 kV constant potential,
molybdenum anode; filtration: 0.03 mm Mo, anode angle 12

,
behind 2 mm thick PMMA compression plate; HVL
1
: 0.32 mm
Al. Reproduced with permission from Dr. J. Th. M. Jansen.
International Commission on Radiation Units and Measurements 2005
tungsten K-edge, but the spectrum is dominated by
the bremsstrahlung contribution.
In x-ray mammography, a spectrum of much
lower energy, for example, from a molybdenum
target, is used to obtain good soft-tissue contrast.
This spectrum is modified by a K-edge filter (gener-
ally molybdenum although rhodium is used in some
situations) and shows strong peaks corresponding to
the K
a
and K
b
characteristic radiation, which for the
molybdenum target have average energies of 17.4
and 19.6 keV, respectively (Figure 2.1). The K-edge
filter reduces the low-energy end of the bremsstrahl-
ung spectrum and much of the bremsstrahlung
spectrum at energies above the K-edge, which for
molybdenum is at 20 keV.
A combination of calculations and measurements
has been used to produce catalogues of x-ray spectra
for specified values of tube voltage, filtration, target
material, target angle, and waveform (Birch et al.,
1979; Seelentag et al., 1979; Iles, 1987). More
recently, the Institute of Physics and Engineering
in Medicine (IPEM) prepared a catalogue of dia-
gnostic x-ray spectra and other data available on
CD-ROM (IPEM, 1997). This allows calculation of
x-ray spectra based upon the method of Birch et al.
(1979) in cases where an appropriate spectrum can-
not be found in one of the available catalogues. Such
information on the spectral distribution of the
photon fluence is required when calculating the
response of detectors or the air kerma or absorbed
dose in different materials when exposed to the
relatively broad photon spectra typical of x rays
used in medical x-ray imaging. It is important to
check calculated values for first and second HVL
against measurements.
2.2 HVL
Routinely, the practical determination of x-ray
beam quality relies on simple attenuation measure-
ments, usually in aluminium, to determine the HVL.
The first, HVL
1
, is the thickness of a specified mater-
ial, which attenuates the beam of radiation to an
extent such that the radiation quantity is reduced
to half its initial value (ICRU, 1970). The use of
different radiation quantities such as exposure or
absorbed dose will lead to different HVL
1
values.
For the characterization of x-ray beams used for
medical imaging, the air kerma, K
a
, or the air
kerma rate,
_
KK
a
, is recommended for the determina-
tion of the HVL. In the definition of HVL the con-
tribution of all scattered radiation, other than any
that might be present initially in the beam con-
cerned (see Section 2.4), is to be excluded. The HVL
alone is often not an adequate specification of the
x-ray beam quality because markedly different
spectra can sometimes result in the same value of
HVL
1
, as illustrated by Figure 2.2. It should be noted
that two of the spectra shown have very low filtra-
tion and would not be legal for use in medical x-ray
imaging in many countries. In the figure, photon
spectra are shown for four x-ray beams having sim-
ilar values of HVL
1
, but generated at different tube
voltages and having different filtration. The x-ray
spectra are rather different and may cause different
dosimeter responses and different dose distributions
in an irradiated medium, for example, phantom or
patient.
For measurement of the HVL the recommenda-
tions of the ICRU (1964) should be followed. It has
been shown that a narrow beam and a sufficiently
large distance between the absorber and measuring
device should be used to obtain the correct HVL.
The instrument used for attenuation measurements
should have weak energy dependence over the
range concerned. The use of a monitor is advisable
to facilitate a correction for variations in the output
of the x-ray tube. The monitor should be positioned
so that its readings are independent of the thickness
of the absorber. By limiting the field diameter, the
amount of scattered radiation recorded will be
reduced, but the field dimensions must be larger
than the sensitive volume of the measuring device.
The collimator must be of sufficient thickness to
absorb the primary beam. A radiographic method
may be used to check the alignment.
The variation of the measured HVL with field
diameter and with detectorabsorber distance is
presented elsewhere (ICRU, 1964). As a general
rule the absorbing material should be placed approx-
imately midway between the source and the
detector, the minimum sourcedetector distance is
50 cm at a maximum field diameter of 5 cm. If
greater accuracy is required, the method described
by Trout et al. (1960) should be used for which a
unique value can be determined by measuring the
HVL for three collimator sizes. A linear extrapola-
tion of a plot of HVL against the field diameter will
yield the zero-field-area value.
2.3 X-RAY TUBE VOLTAGE
The x-ray tube voltage may be measured using
either invasive or non-invasive equipment. Both
approaches have advantages and disadvantages.
Invasive measurements can be made, for example,
by means of a frequency-compensated high-voltage
divider combined with a storage oscilloscope. The
calibration of the high-voltage divider should ideally
be traceable to a primary or secondary voltage
standard. For connection to the oscilloscope, high-
impedance probes and low-impedance cables have
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
22
to be used. The oscilloscope probes should preferably
be calibrated in conjunction with the oscilloscope,
for the range of voltages to be measured.
In the past, the x-ray peak tube voltage has been
checked with an uncertainty of 12 kV using a modi-
fied Ardran and Crookes test cassette (Jacobson
et al., 1976). Although the test film produced with
this cassette can be evaluated visually, reliable res-
ults can only be obtained by using a densitometer
to read the patches on the film.
Non-invasive electronic devices for measurement
of the x-ray tube voltage are generally based on
attenuation measurements (Gard, 1996) but spectro-
metric methods may also be used. Methods differ as
to which value is indicated, for example, peak volt-
age or average voltage. Different instruments are
usually required for general radiology and mammo-
graphy. They need careful calibration against a
primary or secondary standard. The measured
value of the x-ray tube voltage will be influenced by
the degree of beam filtration and may be dependent
on the air kerma rate. Devices that rely on the
integrated or the peak signal cannot measure or
indicate the waveform, and it is desirable to addi-
tionally use an oscilloscope or similar device.
For both invasive and non-invasive devices
calibration facilities are essential, but primary or
secondary standard devices are not available in all
countries. While the quality of the calibration inev-
itably plays an important role, it must be appreci-
ated that the vagueness of the term peak voltage is
also a contributory factor (Kramer et al., 1998).
Therefore, a new quantity termed the practical
peak voltage has been proposed (IEC, 1996; Kramer
et al., 1998). This quantity is based on the concept
that the radiation generated by a high voltage of
any waveform produces the same contrast as radi-
ation generated by an equivalent constant potential
generator. The constant potential producing the
same contrast for a specified contrast configuration
and specified x-ray tube properties as the waveform
under test is the practical peak voltage. This quant-
ity is derived from the contrast produced by a
reference x-ray tube irradiating a 10 cm thick
0 20 40 60 80 100 120
PHOTON ENERGY; E/keV
R
E
L
A
T
I
V
E
N
U
M
B
E
R
O
F
P
H
O
T
O
N
S
0 20 40 60 80 100 120
PHOTON ENERGY; E/keV
R
E
L
A
T
I
V
E
N
U
M
B
E
R
O
F
P
H
O
T
O
N
S
0 20 40 60 80 100 120
PHOTON ENERGY; E/keV
R
E
L
A
T
I
V
E
N
U
M
B
E
R
O
F
P
H
O
T
O
N
S
0 20 40 60 80 100 120
PHOTON ENERGY; E/keV
R
E
L
A
T
I
V
E
N
U
M
B
E
R
O
F
P
H
O
T
O
N
S
Figure 2.2. Calculated spectra (IPEM, 1997) of filtered x-ray beams with almost the same HVL
1
but generated at different tube voltages
in a tungsten anode (constant potential, anode angle 16

) and with different filtration. Top left tube voltage, 60 kV; filtration, 4.3 mm
Al; HVL
1
, 2.74 mm Al; HVL
2
, 3.5 mm Al. Top right tube voltage, 75 kV; filtration, 2.9 mm Al; HVL
1
, 2.74 mm Al; HVL
2
, 3.9 mm Al;
Bottom left tube voltage, 90 kV; filtration, 2.0 mm Al; HVL
1
, 2.72 mm Al; HVL
2
, 4.3 mm Al. Bottom right tube voltage, 110 kV;
filtration, 1.2 mm Al; HVL
1
, 2.68 mm Al; HVL
2
, 4.8 mm Al. Reproduced with permission from Dr. J. Th. M. Jansen.
SPECIFICATION OF X-RAY BEAMS
23
PMMA phantom covered partially with a piece of
1 mm Al as a contrast medium (Kramer et al., 1998).
The result of the contrast-equivalent peak voltage
is insensitive to small differences in the contrast
geometry.
The experimental results of Baorong et al. (2000)
demonstrated that a contrast-equivalent x-ray tube
voltage can be determined with an accuracy ranging
from 200 V to 3 kV, depending on the magnitude
of the tube voltage in the range of 40150 kV. The
reference contrast geometry of 1 mm Al on 10 cm
PMMA selected on the basis of calculations (Kramer
et al., 1998) also appears suitable from experimental
evidence.
2.4 TOTAL FILTRATION
The materials of the x-ray tube window and any
permanent filters will attenuate the x rays produced
in the target (focal spot) of the x-ray tube. The thick-
ness of these materials is called the inherent filtra-
tion and often expressed in an equivalent thickness
of aluminium. There may be additional filtration
fromthe radiolucent mirror of the light-beamsystem
and from other parts of the collimator system.
Often additional (aluminium) filtration is applied
to remove the lower-energy photons, which will con-
tribute only to patient dose and not to the image. The
total filtration is the sum of the inherent and the
additional filtration.
The HVL can be used to estimate the total
filtration of x-ray tubes used in medical imaging.
The uncertainty of the total filtration will be within
10 % (Gilmore and Cranley, 1990) provided that the
HVL
1
can be measured within 0.06 mm Al, the tube
voltage within 1.9 kV, the tube voltage ripple within
5 %, and the effective target angle is known within
1.8

. This procedure presumes the availability of


accurate data relating HVL
1
to tube total filtration
such as presented in IPEM Reports 64 (1991) and 78
(1997).
2.5 X-RAY TUBE OUTPUT
The x-ray tube output may be used in conjunction
with the inverse-square law to calculate the air
kerma incident on a patient or a phantom if the
tube-current exposuretime product is known
(Section 3.2.1). The x-ray tube output, Y(d), is
defined as the quotient of the air kerma, K
a
(d), at a
specified distance, d, from the x-ray tube focal spot
(usually 1 m) by the tube-current exposuretime
product, P
It
. Thus
Y d K
a
d =P
It
: 2:1
Unit: J/(kg/C), or Gy/mAs.
The magnitude of the x-ray tube output will
depend upon the design of the tube, tube voltage,
and filtration and may change as the tube ages.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
24
Journal of the ICRU Vol 5 No 2 (2005) Report 74 doi:10.1093/jicru/ndi025
Oxford University Press
3 QUANTITIES AND UNITS FOR MEASUREMENT AND
CALCULATION IN MEDICAL X-RAY IMAGING
In the first part of this section, basic dosimetric
quantities relevant for medical x-ray imaging are
introduced. The following subsection presents
quantities used for specific applications such as
radiography, fluoroscopy, and CT. Risk-related
quantities are given in Section 3.3. Dose-conversion
coefficients relating mean organ doses, or absorbed
dose to a localized region of tissue, to readily
measurable dosimetric quantities are presented in
Section 3.4. Quantities recommended for the estab-
lishment and use of diagnostic reference levels are
given in the final section.
3.1 BASIC DOSIMETRIC QUANTITIES
The energy carried by the photons in an x-ray
beam can be specified in terms of the energy fluence.
The energy fluence, C, is the quotient of dR by da,
where dR is the radiant energy incident on a sphere
of cross-sectional area da (ICRU, 1998c)
C
dR
da
: 3:1
Unit: J/m
2
.
The conversion of energy refers to the transfer of
energy from ionizing particles to secondary ionizing
particles. The quantity kerma relates to the kinetic
energy of all the charged particles liberated by
uncharged particles; the energy expended against
the binding energies, usually a relatively small
component, is, by definition, not included.
The kerma, K, is the quotient of dE
tr
by dm, where
dE
tr
is the sum of the initial kinetic energies of all
the charged particles liberated by uncharged parti-
cles in a mass dm of material (ICRU, 1998c), thus
K
dE
tr
dm
: 3:2
Unit: J/kg, special name gray (Gy).
Kerma is defined at a point and concerns the ini-
tial transfer of energy by uncharged particles to
matter. For uncharged particles of a single energy,
kerma is related to the energy fluence by the mass
energy transfer coefficient, m
tr
/r of the material for
these particles
K
Cm
tr
r
: 3:3
For x rays used for medical imaging the kerma is
usually expressed in air. The air kerma, K
a
, is
related to the energy fluence by the mass energy
transfer coefficient for air, (m
tr
/r)
a
. Thus for photons
of a single energy, K
a
is given by
K
a
C m
tr
/r
a
: 3:4
Unit: J/kg, special name gray (Gy).
When the photons are not monoenergetic, which is
usually the case in medical x-ray imaging, a mean
value of (m
tr
/r)
a
should be used, weighted according
to the energy distribution of the energy fluence.
The kerma rate,
_
KK, is the quotient of dK by dt,
where dK is the increment of kerma in the time
interval dt, thus
_
KK
dK
dt
: 3:5
Unit: (J/kg)/s, or Gy/s.
The absorbed dose, D, can be used to quantify the
deposition of energy by ionizing radiation. It is
defined in ICRU Report 60 (1998c) as the quotient
of d by dm, where d is the mean energy imparted to
matter of mass dm, thus
D
d
dm
: 3:6
Unit: J/kg, special name gray (Gy).
Under conditions of charged particle equilibrium,
the absorbed dose to material t is related to the
energy fluence by the mass energy absorption coeffi-
cient in that material (m
en
/r)
t
. Thus for photons of a
single energy D
t
is given by
D
t
C m
en
=r
t
: 3:7
Again, when the photons are not monoenergetic, a
mean value of (m
en
/r)
t
should be used, weighted
according to the energy distribution of the energy
fluence.
All the energy that has been transferred from
the ionizing photons to the kinetic energy of the
International Commission on Radiation Units and Measurements 2005
secondary electrons is not necessarily absorbed loc-
ally by the irradiated material as the electrons slow
down. A small fraction gg of the electron energy is
converted directly into photon energy (mostly
bremsstrahlung), but this fraction is negligible for
materials of low atomic numbers Z and for the elec-
tron energies associated with medical x-ray imaging
[for example, gg 6:6 10
4
for 100 keV electrons in
air (ICRU, 1984)]. Strictly,
m
en
r

1 gg m
tr
r
: 3:8
The transfer of energy (kerma) takes place at a
point whereas the subsequent deposition of energy
to matter (absorbed dose) is spread over distances
determined by the range of the secondary electrons.
If gg is negligible and charged-particle equilibrium
exists, i.e., the number, energy, and direction of
the electrons entering a volume of interest are the
same as for those leaving it, then the kerma will be
numerically equal to the absorbed dose, when
expressed in the same material. In medical x-ray
imaging, charged-particle equilibrium is readily
achieved in low atomic number materials such as
soft tissue, since the range of the secondary electrons
is very short compared with the mean free path of
the x-ray photons, which produced them. As a con-
sequence, the photon fluence changes imperceptibly
over the distance in which the electron energy is
imparted to the medium.
Discontinuities in charged-particle fluence caused
by out-scattered electrons of low energy will occur at
locations of rapid density or Z changes such as
boundaries between soft and bony tissues. For x
rays used in medical imaging, a fluence enhance-
ment by a factor as large as 100 at the media
interface should be accounted for, decaying exponen-
tially over a range of some mm or tens of mm (Regulla
et al., 1998).
For colon examinations, the influence of the con-
trast modification through a barium (29 % weight/
volume barium sulfate) meal or by insufflating air
has been investigated for a limited number of pro-
jections (Schultz et al., 2001). Inflation of the intest-
ines with air has little influence on the radiation
burden, i.e., it yields about 1 % higher value for the
dose-conversion coefficient of the examination com-
pared to the normal situation. The barium meal
decreases the dose-conversion coefficient of the
examination by up to about 18 %. Ignoring this
decrease results in a slight overestimation of the
dose, thus being safe with respect to radiation pro-
tection of the patient.
In the case of charged-particle equilibrium and in
the absence of bremsstrahlung losses, the kerma is
numerically equal to the absorbed dose. Thus
K D C m
en
=r C m
tr
=r : 3:9
Despite this equivalence of kerma and absorbed
dose, it is the air kerma measured free-in-air on the
central axis of the x-ray beam at a specified distance
from the focal spot that is the most common method
for specifying the output of x-ray tubes used for
medical imaging. For x-rays used in medical imag-
ing, calibrations are made in terms of air kerma
because this is the dosimetric quantity supported
by the international metrology system and for
which primary standards have been developed.
The absorbed dose rate,
_
DD, is defined by the ICRU
(1998c) as the quotient of dD by dt, where dD is
the increment of absorbed dose in the time interval
dt, thus
_
DD
dD
dt
: 3:10
Unit: (J/kg)/s, or Gy/s
The exposure, X, is defined by the ICRU (1998c) as
the quotient of dQ by dm, where dQ is the absolute
value of the total charge of the ions of one sign pro-
duced in air when all the electrons and positrons
liberated or created by photons in air of mass dm
are completely stopped in air, thus
X
dQ
dm
: 3:11
Unit: C/kg.
The exposure relates to the first stage of the inter-
action process in the same way as air kerma. Expos-
ure is essentially the ionization equivalent of air
kerma. When bremsstrahlung losses are negligible,
the relationship between them is given by
K
a

XW
a
e
, 3:12
where W
a
is the mean energy required to produce an
ion pair in air and e is the elementary charge. The
establishment of complete charged-particle equilib-
rium (which may be difficult to achieve to the level
needed for accurate calibration, particularly for
higher photon energies) is not required.
3.2 APPLICATION-SPECIFIC QUANTITIES
Several practical dosimetric quantities have been
found useful for measurements in medical x-ray
imaging. However, ambiguity exists in names of
quantities and their use (Section 1.2). Owing to
the equivalence of numerical values of absorbed
dose and kerma in the same material for the x-ray
energies used in medical imaging, quantities have
often been alternatively referred to in terms of
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
26
absorbed dose (usually abbreviated to dose) or in
terms of kerma. Commonly used names for these
quantities are
It is necessary to specify the position of the point of
measurement or calculation of the quantities with
respect to the x-ray tube focal spot and the patient
or phantom. Because diverging radiation beams are
invariably used in medical imaging, the kerma
and dose will decrease with distance from the
x-ray tube focal spot approximately in accordance
with the inverse-square law. This is in contrast to
the situation modelled in some occupational-
dosimetry calculations where broad unidirectional
beams are assumed and the air kerma is constant
for all points in the radiation field incident on the
body. Radiation backscattered from within the
patient or a phantom representing the patient will
make a significant contribution to the kerma or dose
at the entrance surface; backscatter factors range
from 1.25 to 1.60 for general radiology (Petoussi-
Henss et al., 1998). Consequently, the distance of
the point of measurement from the patient or phan-
tom needs to be specified.
The first four (pairs of ) kerma (or dose) quantities
listed above refer to the same position (that is,
the point where the central axis of the x-ray beam
intercepts the plane corresponding to the entrance
surface of the patient or phantom, Figure 3.1). How-
ever, the first two pairs are to be determined free-in-
air, i.e., in the absence of the patient or phantomand
the third pair is to be determined in the presence of
the patient or phantom.
Because almost all these quantities will be meas-
ured with instruments calibrated in terms of air
kerma, it is considered to be appropriate and more
consistent with practice elsewhere (Section 1.2) to
focal spot-
to-surface
distance
d
FSD
focal spot-
to-image receptor
distance
d
FiD
focal spot
position
collimator
x-ray tube
air kerma area product P
KA
- incident air kerma (no backscatter) K
a,i
- entrance surface air kerma
(including backscatter)
K
a,e
absorbed dose to tissue at a point in the patient D
t
table
organ dose D
T
meter
image receptor
Figure 3.1. Simple exposure arrangement for radiography showing some of the dosimetric and geometric quantities recommended in the
present Report for determination of patient dose.
Exposure at skin entrance (free-in-air)
Exposure rate at skin entrance (free-in-air)
Entrance surface kerma
(or dose)
(free-in-air)
Entrance surface kerma
(or dose) rate
(free-in-air)
Entrance surface dose
(or kerma)
(with backscatter)
Entrance surface dose
(or kerma) rate
(with backscatter)
Kerma (or dose)area product (free-in-air)
Kerma (or dose)area
product rate
(free-in-air)
CT dose (or kerma) index (free-in-air or
in phantom)
MEDICAL X-RAY IMAGING MEASURING UNITS AND CALCULATION
27
name them in terms of air kerma, except when meas-
ured or calculated inside a phantom or a patient. In
the latter case absorbed dose is the preferred quant-
ity. Additional qualifying words are used to indicate
the position of measurement and whether backs-
cattered radiation from the patient is to be included
or not. To indicate whether backscatter is included
or not in the air kerma the terms incident (no backs-
catter) and entrance surface (including backscatter)
are used. For the first four quantities, subscripts are
added to the symbol for the quantity. The first sub-
script specifies the material in which the quantity
is expressed, for example, air indicated by a.
The second subscript indicates the measurement
condition i.e., incident or entrance surface quantity
indicated by i or e. Thus incident air kerma and
entrance surface air kerma are denoted as K
a,i
and
K
a,e
, respectively.
The air kermaarea product is the integral of
the air kerma over the area of the x-ray beam in a
plane perpendicular to the beam axis. If the air
kerma is constant over the beam area, the integral
becomes equal to the product of the air kerma and
the area, hence the name air kermaarea product.
The symbol recommended in this Report for the
air kermaarea product is P
KA
. The symbol P indic-
ates that the quantity is a product and the subscript
KA indicates that the factors in the product are
the air kerma and area. The air kermaarea
product rate,
_
PP
KA
, is defined as the quotient of the
increment in the air kermaarea product by the time
interval dt.
Like the air kermaarea product, the air
kermalength product can be defined as the integral
of the air kerma along a line. This quantity is useful
in CT where the line is chosen to be parallel to the
axis of rotation of the CT scanner. If the air kerma is
constant over a length L and equal to zero elsewhere
along the line, the integral becomes equal to the
product of the air kerma and the length, hence the
name air kermalength product. The symbol recom-
mended in the present Report for the air kerma
length product is P
KL
. The symbol P indicates that
the quantity is a product and the subscript KL
indicates that the factors in the product are the air
kerma and length. Similarly the doselength prod-
uct, P
DL
, can be defined.
Also for dosimetry in CT, the CT air kerma index
(symbol C
K
) for measurements free-in-air can be
defined as the air kermalength product, P
KL
divided by the nominal slice thickness, T or the
sum of slice thicknesses in the case of a multiple
slice scanner. As the nominal slice thickness is
used (rather than the actual slice thickness), the
symbol C
K
is used instead of air kerma with a relev-
ant subscript.
Absorbed-dose measurements in CT are often
made in special head and body dosimetry phantoms
(ICRU, 1992c) and the CTDI (symbol C
D
) is used,
which is defined similarly to the CT air kerma
index, as the doselength product, P
DL
, divided
by the nominal slice thickness T. It should be
stressed, however, that problems occur for measure-
ment of the dose to air because the phantommaterial
(PMMA) and the material in which the dose is
expressed (air) differ (Section 4.2.5). Therefore, the
CT air kerma index in-phantom(symbol C
K,p
) should
be used instead of the CTDI and, similarly, the air
kermalength product in-phantom (symbol P
KL,p
)
instead of the doselength product.
Recommended names, symbols, and field of
application of these quantities are given in
Table 3.1. More detailed definitions are given
below.
3.2.1 Incident air kerma and incident
air kerma rate
The incident air kerma is the air kerma from the
incident beam on the central x-ray beam axis at the
focal spot-to-surface distance, d
FSD
, i.e., at the skin-
entrance plane (Figure 3.1). Only the primary radi-
ation incident on the patient or phantom and not the
backscattered radiation, is included. In the present
Report it is given the symbol K
a,i
.
Unit: J/kg, special name: gray (Gy).
The incident air kerma is approximately related to
the air-kerma free-in-air at any other distance, d,
fromthe tube focal spot, K
a
(d), by the inverse-square
law. Thus
K
a;i
K
a
d d=d
FSD

2
: 3:13
Table 3.1. Recommended application specific quantities
for dosimetry in medical x-ray imaging
a
.
Quantity name Symbol Field of application
Incident air kerma K
a,i
Radiography and
fluoroscopy
Incident air kerma rate
_
KK
a;i
Fluoroscopy
Entrance surface air kerma K
a,e
Radiography and
fluoroscopy
Entrance surface air kerma rate
_
KK
a;e
Fluoroscopy
Air kermaarea-product P
KA
Radiography and
fluoroscopy
Air kermaarea-product rate
_
PP
KA
Radiography and
fluoroscopy
Air kermalength product P
KL
CT
CT air kerma index C
K
CT
a
All quantities are used to assess stochastic effects. Quantities
K
a,i
and K
a,e
are also useful to monitor the maximum skin dose,
also referred to as the peak skin dose (Miller et al., 2002), for
deterministic effects in interventional radiology.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
28
It is stated approximately because there are
several small corrections due to attenuation in air,
scatter in air, and x-ray source structures. The incid-
ent air kerma can be easily calculated from the x-ray
tube output, Y(d) (Section 2.5), provided the d
FSD
and the tube-current exposuretime product are
known for the specified radiation quality.
The incident air kerma rate,
_
KK
a;i
, is the quotient of
dK
a,i
by dt, where dK
a,i
is the increment of incident
air kerma in the time interval dt, thus
_
KK
a;i

dK
a;i
dt
: 3:14
Unit: J/(kg/s), or Gy/s.
3.2.2 Entrance-surface air kerma and
entrance-surface air kerma rate
The entrance-surface air kerma is the air kerma
on the central x-ray beam axis at the point where
the x-ray beam enters the patient or phantom
(Figure 3.1). The contribution of backscattered
radiation is included. In the present Report it is
given the symbol K
a,e
.
Unit: J/kg, special name: gray (Gy).
The entrance-surface air kerma is related to
the incident air kerma by the backscatter factor,
B. Thus
K
a;e
K
a;i
B: 3:15
The backscatter factor depends on the x-ray spec-
trum, the x-ray field size, and the thickness and
composition of the patient or phantom. Additional
information on back scatter factors, including tables,
is given in Appendix A.
The entrance-surface air-kerma rate,
_
KK
a;e
, is the
quotient of dK
a,e
by dt, where dK
a,e
is the increment
of entrance surface air kerma in the time interval dt,
thus
_
KK
a;e

dK
a;e
dt
: 3:16
Unit: J/(kg/s), or Gy/s.
3.2.3 Air kermaarea product and
air kermaarea product rate
The air kermaarea product is the integral of the
air kerma free-in-air over the area A of the x-ray
beam in a plane perpendicular to the beam axis. In
the present Report it is given the symbol P
KA
. Thus
P
KA

Z
A
K
a
A dA: 3:17
Unit: J/(kg/m
2
), or Gy cm
2
If the air kerma free-in-air K
a
(A) is constant over
the beam area, which is approximately valid for
small beam areas,
P
KA

Z
A
K
a
A dA K
a
A: 3:18
The P
KA
has the useful property of being approx-
imately (when air attenuation and scatter and extra
focal irradiation can be neglected) invariant with
distance from the x-ray tube focal spot, as long as
the plane of measurement or calculation is not so
close to the patient or phantom as to receive a signi-
ficant contribution from backscattered radiation.
Usually, the position of the plane does not need to
be specified.
The air kermaarea product rate,
_
PP
KA
is the quo-
tient of dP
KA
by dt, where dP
KA
is the increment of the
air kermaarea product in the time interval dt, thus
_
PP
KA
dP
KA
=dt: 3:19
Unit: Jm
2
/(kg s), or Gy cm
2
/s.
3.2.4 Air kermalength product
The air kermalength product is the integral of the
air kerma free-in-air over a line L parallel to the axis
of rotation of a CT scanner. In the present Report it
is given the symbol P
KL
. Thus
P
KL

Z
L
K
a
L dL: 3:20
Unit: (J/kg) m, or Gy cm.
If the air kerma free-in-air K
a
(L) is constant over a
length L and equal to zero elsewhere,
P
KL

Z
L
K
a
L dL K
a
L: 3:21
For measurements inside the standard CT dosi-
metry phantoms, the air kermalength product in-
phantom, P
KL,p
, can be defined similarly to P
KL
.
3.2.5 CT air-kerma index free-in-air
This quantity relates particularly to the fan-
shaped beams used in CT. The CT air kerma index
is the integral of the CT axial air kerma profile,
K
a
(z), along the axis of rotation of the CT scanner, z
for a single rotation with a single slice divided by the
nominal slice thickness T. In the present Report it is
given the symbol C
K
,
C
K

1
T
Z
1
1
K
a
z dz
P
KL
T
: 3:22
Unit: J/kg, or Gy.
For a multi-slice scanner with N
i
slices of thick-
ness T
i
, C
K
becomes
C
K

P
KL
N
i
T
i
: 3:23
MEDICAL X-RAY IMAGING MEASURING UNITS AND CALCULATION
29
3.2.6 CT air-kerma index in the standard
CT dosimetry phantoms
For measurements inside the standard PMMA CT
dosimetry phantoms, the CT air-kerma index,
C
K,PMMA
, can be defined similarly to C
K
. The
C
K,PMMA
is the integral of the CT axial air-kerma
profile, K
a,PMMA
(z), inside the PMMA head or body
phantom along an axis, z parallel to the axis of rota-
tion of the CT scanner, for a single rotation divided
by the nominal slice thickness T. In the present
Report it is given the symbol C
K,PMMA
C
K;PMMA

1
T
Z
1
1
K
a;PMMA
z dz
P
KL;PMMA
T
:
3:24
Unit: J/kg, or Gy.
The phantoms are to be placed centrally in the
gantry of the CT scanner with their longitudinal
axis coinciding with the axis of rotation of the scan-
ner. Different approaches have been proposed for the
region of integration in Eq. (3.24) and the material in
which the absorbed dose is used instead of the air
kerma.
The performance standard for CT systems in the
USAspecified an integration distance of 14 times the
nominal slice thickness and PMMA as the phantom
material in which absorbed dose, D
PMMA
, is meas-
ured as well as expressed (Shope et al., 1981; FDA,
1997). The IEC (1999) and EC (2000) recommend an
integration length of 100 mm, which is now also
common practice in the USA. The IEC (1999) allows
for multiple slices of the same thickness in a single
rotation. Both the IEC (1999) and the EC (2000)
recommend that the absorbed dose be expressed in
air, D
a
. Following the IEC and EC approach, the CT
air-kerma index in-phantom for an integration
length of 100 mm, C
K,PMMA,100
, is defined as
C
K;PMMA;100

Z
50 mm
50 mm
K
a;PMMA
z dz
N
i
T
i
, 3:25
where N
i
is the number of slices of thickness T
i
produced in a single rotation. The FDA, IEC, and
EC approaches use the same standard CT dosimetry
phantoms for head and body, respectively (ICRU,
1992c; IEC, 1999).
3.2.7 Weighted CT air-kerma index and
normalized weighted CT air-kerma index
The C
K,PMMA,100
values in the centre
(C
K,PMMA,100,c
) and at the periphery (C
K,PMMA,100,p
)
of the standard CT dosimetry phantoms for head or
body, respectively, may be used to provide an indica-
tion of the average air-kerma in-phantom over a
single rotation (after EC, 2000). For the present
purpose the weighted C
K,PMMA
, denoted by
C
K,PMMA,w
in the standard CT dosimetry phantoms,
is defined (EC, 2000)
C
K;PMMA;w

C
K;PMMA;100;c
2C
K;PMMA;100;p
3
: 3:26
Unit: J/kg, or Gy.
Similar to the European guidelines on quality
criteria for CT (EC, 2000), it is recommended that
C
K,PMMA,100
be determined at the centre
C
K,PMMA,100,c
and at 10 mm below the surface
C
K,PMMA,100,p
of the standard CT dosimetry phan-
toms and that C
K,PMMA,100,p
should represent an
average of measurements at four locations around
the periphery of the phantom.
The subscript n is used to denote the value of
C
K,PMMA,w
that has been normalized to unit tube-
current exposuretime product, P
It
(after EC, 2000)
n
C
K;PMMA;w

C
K;PMMA;w
P
It
, 3:27
where P
It
is the tube-current exposuretime
product (unit: mAs).
Unit: J/(kg/C), or Gy/(mAs).
3.2.8 CT air kermalength product
The CT air kermalength product in the standard
head or body CT dosimetry phantom for a complete
conventional (serial CT) examination is defined as
P
KL;CT

X
j
n
C
K;PMMA;w
j
T
j
N
j
P
It
j
, 3:28
where j represents each serial scan sequence form-
ing part of an examination and N
j
is the number of
slices, each of thickness T
j
(cm) and tube-current
exposuretime product P
It
j
(mAs), in a particular
sequence.
Unit: J/kg, or Gy.
Any variations in applied tube voltage setting dur-
ing the examination will require corresponding
changes in the value of
n
C
K,PMMA,w
i
used.
In the case of helical (spiral) scanning
P
KL;CT

X
i
n
C
K;PMMA;wi
T
i
I
i
t
i
, 3:29
where, for each of i helical sequences forming part of
an examination, T
i
is the nominal irradiated slice
thickness (cm), I
i
is the tube current (mA), and t
i
is
the total acquisition time (s) for the sequence.
Quantity
n
C
K,PMMA,w
i
is determined for a single rota-
tion as in serial scanning.
3.3 RISK-RELATED QUANTITIES
The fundamental dosimetric quantity in radiolo-
gical protection is the absorbed dose D. At low dose
levels, the mean absorbed doses in organs or tissues
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
30
in the human body are taken to be indicators of the
probability of subsequent stochastic effects; at high
dose levels, absorbed doses to the more heavily irra-
diated sites within the body are taken to be indicat-
ors of the severity of deterministic effects (ICRP,
1991b, 2000). In view of the obvious difficulties in
measuring the distribution of absorbed dose within
the body during medical x-ray imaging, more prac-
tical dosimetric quantities have evolved which can
be directly measured or readily estimated from clo-
sely related measurements. The risk-related quant-
ities can be obtained from the practical dosimetric
quantities using the dose-conversion coefficients in
Section 3.4.
3.3.1 Absorbed dose in relation to
deterministic effects
The threshold doses for deterministic effects are
not approached in most applications in diagnostic
radiology. However, this is not the case for, e.g.,
interventional radiology where cases of erythema,
more serious skin injuries and epilation have been
reported following extensive periods of fluoroscopic
imaging (ICRP, 2000). If deterministic effects are
considered to be a possibility, the absorbed dose to
the more heavily irradiated regions of tissues at the
surface of the body is the radiation quantity of inter-
est. The more heavily irradiated tissues will usually
comprise those localized areas of the skin that lie
within the primary x-ray beam for the longest time
during these interventional procedures. The mean
absorbed dose D
T,local
or D
skin,local
(see also the fol-
lowing section) in the localized region of tissue of
interest is then the required quantity, where the
subscript skin,local denotes the tissue region in
which the absorbed dose is determined, also the
material in which the dose is expressed should be
indicated. The present quantity is also referred to as
peak skin dose (Miller et al., 2002).
3.3.2 Absorbed dose for assessment of
stochastic effects (organ dose)
Absorbed dose, D is defined in Section 3.1 at a
point. The ICRP has recommended that the appro-
priate dosimetric indicator for the probability of
stochastic radiation effects is the average absorbed
dose in a tissue or organ (ICRP, 1991b). In ICRU
Report 51 (1993b) the mean absorbed dose in a spe-
cified organ or tissue T has been given the symbol D
T
and is defined either as the integral of the absorbed
dose D
t
over the mass of the tissue divided by its
mass m
T
D
T

1
m
T
Z
m
T
D
t
dm, 3:30
where D
t
is the absorbed dose at a point in tissue
material t, or as the energy imparted (ICRU, 1998c)
to the tissue,
T
, divided by its mass
D
T


T
m
T
: 3:31
The mean absorbed dose in a specified organ or
tissue is further simply referred to as organ dose.
The subscript T can be replaced by a specific organ,
for example, stomach, D
stomach
. In mammography,
the mean absorbed dose to the radiosensitive tissues
in the breast, i.e., the glandular tissues, provides a
sufficient measure for assessment of the radiation
risk (ICRP, 1987).
3.3.3 Equivalent dose and effective dose
The probability of stochastic events is found to
depend not only on the absorbed dose but also on
the type and energy of the radiation depositing the
dose (ICRP, 1991b). The ICRP has, therefore, recom-
mended that for radiological protection purposes,
organ dose should be weighted for radiation quality
by a radiation weighting factor, w
R
. Values of w
R
have been selected by the ICRP to be representative
of the relative biological effectiveness of specified
types and energies of radiation incident on the sur-
face of the body in inducing stochastic effects at low
doses (ICRP, 1991b). The present weighted absorbed
dose has been given the name equivalent dose and
the symbol H
T
. The special name of sievert (symbol
Sv) has been given to the unit of equivalent dose;
1 Sv 1 J/kg. The ICRP uses a radiation weighting
factor of unity (1 Sv/Gy) for photons of all energies,
so for the x rays used in medical imaging organ dose,
D
T
, expressed in Gy is numerically equal to equival-
ent dose, H
T
, expressed in Sv.
The radiation weighting factor is intended to do no
more than to provide a rough indication of the
changes in the biological effectiveness with radiation
quality, but it was often interpreted to imply spuri-
ous precision, which the ICRP (1991b) hopes no
longer will be the case. Deviations in radiobiological
effectiveness (RBE) from unity can occur. For
instance, for the low-energy photons employed in
mammography, Brenner and Amols (1989) suggest
an RBE of about 2 compared to higher-energy
gamma rays and of 1.3 compared to 80 kV x rays.
The relationship between the probability of stoch-
astic effects and equivalent dose also depends on the
organ or tissue irradiated. In medical x-ray imaging
more than one organ is often irradiated. It might
therefore be useful to combine the doses to different
tissues in such a way that the combined value is
likely to correlate well with the total of the stochastic
effects. For the radiological protection of workers
MEDICAL X-RAY IMAGING MEASURING UNITS AND CALCULATION
31
and the whole population, the ICRP (1991b) has
defined the factor by which the equivalent dose in a
tissue or organ has to be weighted, called the tissue
weighting factor, w
T.
The tissue weighting factor
represents the relative contribution of that organ
or tissue to the total detriment due to these
effects resulting from uniform irradiation of the
whole body.
The effective dose E (ICRP, 1991b) is defined as
the sum of the weighted equivalent doses in all the
tissues and organs of the body
E
X
T
w
T
H
T
: 3:32
As it is desirable that a uniform equivalent dose to
the whole body should give an effective dose numer-
ically equal to that uniform equivalent dose (ICRP,
1991b), the sum of the tissue weighting factors is
normalized to unity
X
T
w
T
1: 3:33
The values of the tissue weighting factor proposed
by the ICRP are independent of the type and energy
of the radiation incident on the body. These simpli-
fications may be no more than approximations of the
true biological situation (ICRP, 1991b). The con-
sequences following an absorbed dose also depend
on the distribution of the dose in time. The effect of
all exposure conditions other than those dealt with
by the radiation and tissue weighting factors is cov-
ered by using different values of the coefficients
relating equivalent dose and effective dose to the
probability of stochastic effects, rather than by
using additional weighting factors in the definitions
of the quantities (ICRP, 1991b).
The values of both the radiation and tissue weight-
ing factors depend on current knowledge of radiobio-
logy and may change from time to time (ICRP,
1991b). Although such changes are infrequent they
can cause confusion. The definitions of equivalent
dose and effective dose are not confined to any par-
ticular set of these weighting factors. If values of
weighting factors other than those recommended
by the ICRP (1991b) are used, this fact should be
clearly stated and the values should be explicitly
given when the quantities are introduced.
For the estimation of the likely consequences of an
exposure of a known population, it will be sometimes
better to use absorbed dose and specific data relating
to the relative biological effectiveness of the radi-
ations concerned and the probability coefficients
relating to the exposed population (ICRP, 1991b).
In medical x-ray imaging, both the irradiation con-
ditions and the exposed group of patients are in
principle known. Effective dose should not be used,
for example by applying the nominal fatality coeffi-
cients (ICRP, 1991b), for assessment of detriment
from exposure due to medical x-ray imaging. Such
assessments could be inappropriate because of
potential differences in health status, gender, and
age between a particular group of patients and the
reference population for whom the ICRP derived the
risk coefficients. Furthermore, the conditions of low
dose and especially of low dose rate may not always
be met in practice in medical imaging. For an assess-
ment of the risk due to induction of stochastic
and deterministic effects by medical x-ray imaging
detailed knowledge is required of organ doses,
absorbed-dose distribution, and the age and gender
of the group of patients concerned, rather than
effective dose. Therefore, in the present Report
effective dose is not further discussed.
3.4 DOSE-CONVERSION COEFFICIENTS FOR
ASSESSMENT OF ORGAN AND TISSUE DOSES
A conversion coefficient, c, relates a dosimetric
quantity to some other quantity, i.e., the normaliza-
tion quantity, which can be readily measured or cal-
culated in the clinical situation. In general
c
specified dosimetric quantity
normalization quantity
: 3:34
For stochastic effects, the specified dosimetric
quantity is either the mean absorbed dose in an
organ, D
T
or in a specialized tissue of interest, such
as glandular tissue in the breast. For deterministic
effects, the specified dosimetric quantity is the
absorbed dose to the more heavily irradiated regions
of tissues at the surface of the body; the tissue of
interest is usually the localized region of skin that
lies in the primary x-ray beamand receives the high-
est absorbed dose for an interventional procedure. In
all cases, the subscript T, for the general case, may
be replaced with the specific tissue or organ of inter-
est when it is deemed helpful, for example, D
stomach
for the specific organ, D
G
for glandular tissue in the
breast, and D
skin,local
for the skin region in the prim-
ary beam receiving the highest absorbed dose.
For general radiology of adults and children, the
specified dosimetric quantity is the organ dose, D
T
.
The incident air kerma, K
a,i
, the entrance surface air
kerma, K
a,e
, or the air kermaarea product, P
KA
, is
used as a normalization quantity. Thus
c
T;K
a;i

D
T
K
a;i
: 3:35
Unit: (J/kg)/(J/kg), or Gy/Gy, or
c
T;K
a;e

D
T
K
a;e
: 3:36
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
32
Unit: (J/kg)/(J/kg), or Gy/Gy, or
c
T;P
KA

D
T
P
KA
: 3:37
Unit: (J/kg)/(Jcm
2
/kg), or Gy/(Gy/cm
2
).
In mammography, the specified dosimetric quant-
ity is the mean glandular dose, D
G
. Conversion coef-
ficients are available relating the incident air kerma,
K
a,i
to D
G
. Thus
c
G;K
a;i

D
G
K
a;i
: 3:38
Unit: (J/kg)/(J kg), or Gy/Gy.
For CT, when stochastic effects are of interest, the
specified dosimetric quantity is the organ dose, D
T
,
and the CT air kerma index, C
K
, may be used as
normalization quantity. Thus
c
T;C
K

D
T
C
K
: 3:39
Unit: (J/kg)/(J kg), or Gy/Gy.
In relation to deterministic effects that may occur
in interventional radiology, the specified dosimetric
quantity should be the maximum absorbed dose to
the localized region in the primary x-ray beam, for
example, skin region. The measurement of the pre-
sent quantity is complicated, however, because the
skin location of maximum dose is not usually known
a priori. To date, only K
a,i
, has been used as a nor-
malization quantity in conversion coefficients for
deterministic effects in interventional radiology
(Stern et al., 1995a). The dose-conversion coeffi-
cients are given by Eq. (3.36), with D
T
being replaced
by mean absorbed dose to a localized region of tissue
in the primary beam, D
T,local
.
3.5 QUANTITIES RECOMMENDED
FOR ESTABLISHMENT AND USE OF
DIAGNOSTIC REFERENCE LEVELS
Numerous surveys have shown wide variations
in the magnitude of absorbed dose to the patient
for the same type of x-ray procedure performed at
different facilities or even within the same facility
(Adrian Committee, 1960; NEXT, 1978; Wochos
et al., 1979; Burkhardt, 1984; Shrimpton et al.,
1991; EC, 1996c; CRCPD, 2002). This has focussed
attention on the possibility of using reference
values as guidance for the levels that can be
achieved using good radiographic technique and
appropriate x-ray equipment, particularly to
foster elimination of values at the high end of the
distributions. As an aid to the optimization of
absorbed dose to the patient, reference values
(ICRP, 1991b) can be specified. These have been
variously called reference dose values (EC, 1996a),
diagnostic reference levels (ICRP, 1996), or guidance
levels (IAEA, 1996a).
The ICRP (1996) introduced the concept of
diagnostic reference levels (DRLs) for patients. For
diagnostic radiology, the ICRP states that these
levels, which are a form of investigation level,
apply to an easily measured quantity at the surface
of a simple standard phantom or a representative
patient. The ICRP characterizes an investigation
level as a threshold set by operating management
that calls for local investigation (often very simple) if
exceeded. The ICRP recommends that diagnostic
reference level (DRL) values should be selected by
professional medical bodies and reviewed at inter-
vals that represent a compromise between the neces-
sary stability and the long-term changes in the
observed distributions.
In Europe, DRLs are being introduced in legisla-
tion. In the Council Directive 97/43/Euratom (EC,
1997) it is stated that Member States of the Euro-
pean Union shall promote the establishment and use
of diagnostic reference levels for diagnostic x-ray
examinations. EC (1997) states that the appropriate
regulatory authorities within each Member State
are required to ensure that guidance on the estab-
lishment and use of DRLs is available, having regard
to European DRLs where available.
Finally, it is stressed that DRLs are aimed at
the management of patient doses consistent with the
clinical imaging information that is required.
This means that in individual cases, the exceeding
of DRLs may be justified in terms of a clinical
requirement, for example, need for additional
diagnostic information, or the unexpected difficulty
of a procedure.
In practice, DRLs have frequently been set using
the distribution of dosimetric quantities observed in
national surveys, by taking the third quartile value
(75 % of the results are equal to or below this value)
of the observed distribution (Shrimpton et al., 1986;
IPSM, NRPB and CoR, 1992; IAEA, 1996a). Periodic
monitoring of the dosimetric quantities (IPSM,
NRPB and CoR, 1992) in the UK for common con-
ventional x-ray examinations has shown that the
mean and third quartile values of the observed dis-
tributions dropped by about 30 % since an earlier
national survey in the 1980s (Shrimpton et al., 1986;
Hart et al., 1996c).
The present report adopts the term DRL from the
ICRP (1996). Local performance can be checked
against DRLs by periodic measurement as part of a
quality assurance program. The dosimetric quantit-
ies used for DRLs are presented in this section. In
interventional radiology, a combination of various
MEDICAL X-RAY IMAGING MEASURING UNITS AND CALCULATION
33
quantities is considered useful, for example, fluoro-
scopy time, total number of images and air
kermaarea product.
3.5.1 Incident air kerma and
entrance-surface air kerma
Incident air kerma and entrance surface air
kerma, K
a,i
and K
a,e
, for a specified series of patients
or a specified phantom are recommended quantities
for the establishment and use of DRLs in medical
x-ray imaging for simple examinations, i.e., projec-
tion radiography. They give no indication of the
extent of the beam or of the area of the patient
being irradiated. They provide a useful measure for
comparing patient doses for a given projection when
the beam areas and directions, and the radiation
qualities are similar.
Incident air kerma and entrance surface air
kerma, K
a,i
and K
a,e
, are also useful quantities for
interventional radiology when deterministic effects
in skin or other surface tissues are considered. In
this case, they are not used as a DRL, but as a direct
indicator of the maximum absorbed dose in the loc-
alized tissues such as skin, also referred to as peak
skin dose (Miller et al., 2002). The ICRP has found
K
a,i
to be very useful in monitoring, prior to (through
clinical protocols) and during procedures (with in-
place measurement devices), when threshold levels
for deterministic effects are likely to be approached
or exceeded (ICRP, 2000).
3.5.1.1 Mean mammary glandular dose
For mammography, D
G
is the recommended
quantity for the establishment and use of DRLs. It
is not measured directly, but is derived from K
a,i
or
K
a,e
for a specified series of patients or a specified
phantom using conversion coefficients. Values of
conversion coefficients are available as a function of
HVL
1
for compressed breasts of various thickness
and composition as well as for reference phantoms.
It is important to note that these conversion
coefficients are model-dependent. They are dis-
cussed in Appendix E.
3.5.2 Incident air kerma rate and
entrance-surface air kerma rate
Incident air kerma rate and entrance surface air
kerma rate,
_
KK
a;i
and
_
KK
a;e
, for a specified series of
patients or a specified phantom are recommended
quantities for the establishment and use of DRLs in
fluoroscopy. They provide a useful measure for com-
paring patient dose rates for a given projection when
the beam areas and directions and the radiation
qualities are similar.
3.5.3 Air kermaarea product
The air kermaarea product P
KA
for a specified
series of patients or a specified phantom is a
recommended quantity for the establishment and
use of DRLs for general radiography and for
complex procedures involving radiography and
fluoroscopy. Quantity P
KA
includes a measure of
the area of the beam as well as the incident air
kerma. It provides a useful measure for comparing
patient doses for a given procedure when the
beam areas and directions, number of images,
fluoroscopy time, and the radiation qualities are
similar. However, when these parameters are not
reasonably similar, comparison of P
KA
values as
a measure of the resulting organ doses can be
misleading.
3.5.4 CT air kermalength product, P
KL,CT
The CT air kermalength product P
KL,CT
determ-
ined for the standard CT dosimetry phantoms is
proposed by the EC (2000) as a quantity for estab-
lishment and use of DRLs for a complete CT
examination. Since there is only limited experience
with the use of the CT air kermalength product
for establishment and use of DRLs, this quantity
is recommended provisionally in the present
Report.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
34
Journal of the ICRU Vol 5 No 2 (2005) Report 74 doi:10.1093/jicru/ndi026
Oxford University Press
4 MEASUREMENT METHODS
Methods are required for the measurement of nor-
malization quantities used in the dose-conversion
coefficients and of quantities recommended for
establishment and use of DRLs. Measurements of
these quantities rely on the use of (i) ionization
chambers, or (ii) solid-state dosimeters, including
those employing radiothermoluminescence. Calib-
ration procedures and accuracy requirements for
these dosimeters when applied to measure x rays
used in medical imaging are discussed first.
4.1 MEASUREMENT AND QUALITY
ASSURANCE OF DOSIMETERS
The general properties of an instrument type are
characterized during type testing, when the
response to different radiation energies, angles of
incidence, doses, dose rates, and other influencing
parameters are measured. International standards
for type testing is usually applied (IEC, 1997, 2000),
which specify test methods and performance
requirements.
Before use, a given instrument must be calibrated
to determine the relationship, under specified condi-
tions, between the value indicated by the instrument
and the known value of the quantity of interest
(BIPM, IEC, ISO and OIML, 1984). Thereafter re-
calibrations should be made at intervals not exceed-
ing 2 years, and whenever the instrument has been
repaired or its performance is suspect (AAPM, 1992;
Robertson et al., 1992).
The complexity of the calibration procedure
depends on the particular type of instrument and
its intended use. Re-calibration at one radiation
quality, namely, the reference quality, is often suffi-
cient, as it is unlikely that the dependence of the
response on radiation energy would change signific-
antly. Re-calibration at a single or a few points on
the measurement range will usually be sufficient.
However, sometimes it will be important to check
comprehensively the dependence of the response on
the dose rate, particularly if the dosimeter is used at
widely varying dose rates. In the case of ionization
chambers, it is also important to apply a correction
for temperature and pressure at the time of
calibration, unless the chamber is sealed to the
atmosphere or if the electrometer device automatic-
ally corrects for the density of air. If the electrometer
device automatically corrects for the density of air,
the calibration of accuracy of this correction needs
to be checked. Corrections for ion recombination are
usually insignificant but must be considered when
using very high instantaneous dose rates. Correc-
tions for humidity do not have to be applied.
Simple constancy checks and checks of the various
functions are recommended before each use of the
instrument. Constancy checks may be carried out
using a special check source, in a reproducible
source-detector geometry. Standard uncertainty
due to statistical means (type A uncertainty) should
be 3 % at maximum.
The results of calibration are generally expressed
using a calibration coefficient N
N
K
M
, 4:1
where M is the reading of the instrument, corrected
for influence quantities such as temperature and
pressure, in the known calibration field and K is
the value of the desired quantity in the field. The
desired quantity in any unknown field can then be
obtained by multiplying the reading of the instru-
ment in this field by the calibration coefficient N.
For most dosimeters used in medical imaging, the
desired quantity, K of Eq. (4.1), for the calibration is
the air kerma free-in-air, K
a
, in units of gray. The
value obtained for the desired quantity at the time of
calibration must be traceable to the international
measurement system, i.e., it must be related to
appropriate primary standards throughanunbroken
chain of comparisons. The calibration of a field
instrument is performed under well-defined condi-
tions by comparison of the reading of the instrument
with that of a reference instrument. The reference
instrument is usually a secondary standard or a ter-
tiary standard, calibrated by comparison with a
primary standard or a secondary standard, respect-
ively. Requirements for the characteristics and per-
formance of reference and field instruments have
been published by the AAPM (1992).
International Commission on Radiation Units and Measurements 2005
The uncertainty of dose measurements in medical
x-ray imaging, for comparative risk assessments
as well as for quality assurance, should not exceed
7 % (AAPM, 1992). The requirement is given in
terms of the expanded uncertainty
1
using a coverage
factor of 2. For a more detailed discussion of uncer-
tainties see ISO, IEC, OIML, BIPM (1992), and
IAEA (1994, 2000). For a normal distribution of
uncertainties, the use of a coverage factor of 2
corresponds to the confidence level of 95 %. Consid-
ering all steps that contribute to the measurement
uncertainty, this implies that the expanded uncer-
tainty of the calibration should not exceed 5 %
(AAPM, 1992).
Standards laboratories have established well-
defined conditions for the calibration of instruments.
However, the actual radiation field in which the
instrument is used may differ in energy and spatial
distribution from that of the calibration field. Hence,
depending mainly on the instrument design, the cal-
ibration coefficient obtained may not accurately
apply to the radiation field during the actual meas-
urements. Because of this extra uncertainty, special
calibration conditions closely simulating the actual
measurement conditions in diagnostic radiology
have been developed and recommended (IEC, 2004;
Section 4.1.1).
In a few cases the calibration of dosimeters at the
well-defined conditions of the standards laboratory
is neither possible nor feasible. Air kermaarea
product meters, which are part of a defined mechan-
ical set-up of the x-ray equipment, should be calib-
rated in situ. For TLDs, it is more appropriate to
carry out the calibration at the actual field condi-
tions. The calibration of air kermaarea product
meters is discussed in more detail in Section 4.1.2
and that of TLDs in Section 4.1.3.
4.1.1 Calibration of dosimeters in terms of
air kerma free-in-air
For the calibration of dosimeters used in medical
imaging, in terms of air kerma free-in-air, an appro-
priate arrangement of x-ray generator and tube,
collimators, shutter, filters, and monitor chamber is
needed (IAEA, 1994). For the calibration of dosimet-
ers used for medical imaging except mammography,
x-ray tubes with a tungsten target should be
used. For CT, x-ray tubes with a tungsten target
and added copper filtration should be used. For
mammography, x-ray tubes with a molybdenum tar-
get and molybdenum filter should be used.
The radiation quality is a description of the radi-
ation field by a set of identifying characteristics,
such as x-ray tube voltage, total filtration, and
HVL (Section 2.2). For the reasons described above,
special radiation qualities closely simulating the
actual field conditions have been established, for
both attenuated and unattenuated beams in conven-
tional radiology, for CT, and for beams in mammo-
graphy (IEC, 2004). The attenuated beam qualities
are specified to simulate the beam, which has passed
through the patient. It is recommended that these
radiation qualities be used. Examples for radio-
graphy, CT, and mammography are given in
Tables 4.1, 4.2, and 4.3, respectively.
The requirement for traceability of calibrations
with a small uncertainty necessitates the use of the
same radiation qualities in all steps, i.e., for the
measurements and comparisons with the primary
standard and for the measurements and comparisons
with the reference standard (secondary and tertiary
standards). If the reference standard is calibrated
using other radiation qualities, then the uncertainty
of the calibration of a field instrument and the use-
fulness of adopting the special diagnostic qualities
will depend on the quality of the reference standard,
mainly on the constancy of its response with x-ray
energy.
Air kerma rates used during the calibration
should ideally cover the complete range of air
kerma rates stated for the instrument, i.e., the
range within which the instrument will meet its
specified performance. Air kerma rates up to
10 mGy/s, 100 mGy/s, and 500 mGy/s are usually
needed for dose measurements for conventional
diagnostic radiology, mammography, and CT,
respectively.
The expanded uncertainty of the calibration
should be estimated and shown to be within the
requirement of 5 %. An example of the uncertainty
estimation is given in Table 4.4. This example
is referring to a modern primary standards
laboratory.
Guidance on the estimation of the uncertainties
can be found elsewhere (ISO, IEC, OIML, BIPM,
1992; IAEA, 1994). An international code of
practice for patient dosimetry in diagnostic radi-
ology is presently being developed by the IAEA
(Pernicka et al., 2001). It will include practical
details of calibrations at the standards laboratory
and field measurements. Values for the upper
lmits for the uncertianties for calibration of
dosimeters for medical x-ray imaging at Secondary
1
The expanded uncertainty is equal to the combined stand-
ard uncertainty multiplied by the coverage factor. The
combined standard uncertainty is obtained by combining
in quadrature the standard deviations corresponding to
the uncertainties estimated by statistical means (type A
uncertainties) and by non-statistical means (type B uncer-
tainties) cf. Table 4.4.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
36
Standards Dosimetry Laboratories will be recom-
mended by the IAEA.
4.1.2 Calibration of air kermaarea
product meters
The calibration of air kermaarea product meters
should be carried out every time a meter is installed
on a different x-ray set, whenever ionization cham-
bers or electrometers are changed, or at least once a
year if the instrument is permanently installed on
the same equipment (IPSM, NRPB and CoR, 1992).
The response of the meter, i.e., the indicated value
divided by the true P
KA
value at the entrance plane
Table 4.1. Examples of standard radiation qualities for the calibration of dosimeters in general radiography (Reproduced
from IEC, 2004, with permission from IEC)*. An emitting target of tungsten for the x-ray tube is specified. RQR qualities
refer to unattenuated beams and RQA qualities to attenuated beams.
Radiation
quality
Approximate
a
x-ray
tube voltage/kV
Added filtration/mm Al
(simulating
patient thickness)
Nominal first
HVL/mm Al
b
Homogeneity
coefficient
RQR 2 40 1.42 0.81
RQR 3 50 1.78 0.76
RQR 4 60 2.19 0.74
RQR 5
c
70 2.58 0.71
RQR 6 80 3.01 0.69
RQR7 90 3.48 0.68
RQR 8 100 3.97 0.68
RQR 9 120 5.00 0.68
RQR 10 150 6.57 0.72
RQA 2 40 4 2.2
RQA 3 60 10 3.8
RQA 4 60 16 5.4
RQA 5
d
70 21 6.8
RQA 6 80 26 8.2
RQA 7 90 30 9.2
RQA 8 100 34 10.1
RQA 9 120 40 11.6
RQA 10 150 45 13.3
a
Approximate means: adjust the tube voltage.
b
The purity of the aluminium should be at least 99.8 % (ICRU, 1964).
c
This value is generally selected as the reference radiation quality for unattenuated beams for general radiography applications.
d
This value is generally selected as the reference radiation quality for attenuated beams for general radiography applications.
*The author thanks the International Electrotechnical Commission (IEC) for permission to reproduce information from its International
Standard IEC 61267. All such extracts are copyright of IEC, Geneva, Switzerland. All rights reserved. Further information on the IEC is
available from www.iec.ch
Table 4.3. Examples of standard radiation qualities for the
calibration of dosimeters in mammography (Reproduced
from IEC, 2004, with permission from IEC)*. An emitting
target of molybdenum for the x-ray tube is specified. RQR-
M qualities refer to unattenuated beams and RQA-M
qualities to attenuated beams.
Radiation
quality
Approximate
a
x-ray tube
voltage/kV
Added filtration/mm
Al (simulating
patient thickness)
Nominal
first HVL/
mm Al
b
RQR-M1 25 0.28
RQR-M2
c
28 0.31
RQR-M3 30 0.33
RQR-M4 35 0.36
RQA-M1 25 2 0.56
RQA-M2 28 2 0.60
RQA-M3 30 2 0.62
RQA-M4 35 2 0.68
a
Approximate means: adjust the tube voltage.
b
The purity of the aluminium should be at least 99.8 % (ICRU,
1964).
c
This value is generally selected as the reference radiation quality
for mammography (both RQR-M and RQA-M series).
*See Table 4.1.
Table 4.2. Examples of standard radiation qualities for
the calibration of dosimeters in CT (Reproduced fromIEC,
2004, with permission from IEC)*. An emitting target of
tungsten for the x-ray tube is specified.
Radiation
quality
Approximate
a
x-ray tube
voltage/kV
Added filtration/
mm Cu (simulating
patient thickness)
Nominal
first HVL/
mm Al
b
RQT 8 100 0.2 6.9
RQT 9
c
120 0.25 8.4
RQT 10 150 0.3 10.1
a
Approximate means: adjust the tube voltage.
b
The purity of the aluminum should be at least 99.8 % (ICRU,
1964).
c
This value is generally selected as the reference radiation quality
for CT.
*See Table 4.1.
MEASUREMENT METHODS
37
of the patient, will depend on whether the chamber
is installed on an overtable or undertable x-ray
tube. In the latter case the table will attenuate the
x-ray beam before it reaches the patient. Scatter
conditions may also differ from one installation to
another. There may be inhomogeneities in the x-ray
field due to the heel effect and extra focal radiation,
whose magnitude will be equipment-dependent.
Consequently, air kermaarea product meters,
which are part of a defined mechanical set-up in
the equipment, cannot be calibrated at a standards
laboratory but must be calibrated in situ (IPSM,
NRPB and CoR, 1992).
The desired quantity for air kermaarea product
meters is not simply air kerma but the air
kermaarea product, P
KA
in units of Gy cm
2
. The
traceability chain is established through the air
kerma calibration of the reference instrument, as
for simple dosimeters, and through the use of a suf-
ficiently accurate method to measure the beam area.
Calibration of air kermaarea product meters
should be performed with an x-ray field size not
larger than about 10 cm 10 cm to avoid significant
non-uniformity in the dose rate across the beam. The
x-ray tube should be positioned at its customary dis-
tance from the table (Wall, 1989). Calibration meas-
urements should be made at a position 20 cmabove
the table, where a reference dosimeter and a film
cassette can be conveniently positioned to measure
the air kerma and the field area at the same plane
perpendicular to the beam axis. The beam area can
be determined using a densitometer measuring the
area bounded by the line on a suitably exposed film
where the optical density falls to 50 % of its max-
imum value. Examples of calibration procedures
are given by IPSM, NRPB and CoR (1992), and by
Larsson et al. (1996, 1998).
Type testing of air kermaarea product meters
(IEC, 2000) by the manufacturers should establish
that the response is uniform over the entire area of
the ionization chamber and is sufficiently unaffected
by changes in x-ray quality and dose rate. If the
calibration or type testing is performed at a stand-
ards laboratory, the radiation qualities given in
Table 4.1 should be used. The expanded uncertainty
of the calibration should not exceed 5 %, so that the
calibrated instrument is capable of measurements
with the required uncertainty of 7 % for the x-ray
qualities and dose rates encountered in practice.
Experience with the ageing of air kermaarea prod-
uct meters, for example, when installed at interven-
tional radiology systems and submitted to very large
doses during several years, is still limited.
4.1.3 Calibration of TLDs
The sensitivity of TLD systems needs to be
checked regularly by measuring the response to x-
ray exposures of known magnitude and quality. The
magnitude is to be determined in terms of the air
kerma obtained by a reference dosimeter that has
been calibrated in a manner that is traceable to
Table 4.4. Estimated relative uncertainties for the calibration of a dosimeter for medical x-ray imaging using the standard
radiation qualities RQR.
Uncertainty component Type A uncertainty
a
(1 SD) %
Type B uncertainty
b
(1 SD) %
Measurement of air kerma with the reference instrument (secondary or tertiary standard)
Air kerma calibration coefficient of the reference instrument 0.56 0.81
Constancy of the air kerma calibration coefficient 0.29
Reading accuracy 0.50
Differences in calibration spectra 0.58
Correction for temperature and atmospheric pressure 0.03
Correction for ion recombination 0.00
Measurement of air kerma with the instrument to be calibrated
Positioning at the calibration distance 0.06
Non-uniformity of the calibration field 0.06
Correction for temperature and atmospheric pressure 0.03
Reading accuracy 0.50
Square root of the sum of squares 0.90 1.04
Combined standard uncertainty
c
1.38
Expanded uncertainty with coverage factor of 2
d
2.76
a
The uncertainty at one standard deviation estimated by statistical means (ISO, IEC, OIML, BIPM, 1992; IAEA 1994).
b
The uncertainty at one standard deviation estimated by non-statistical means (ISO, IEC, OIML, BIPM, 1992; IAEA 1994).
c
The combination in quadrature of the standard deviations corresponding to type A and type B uncertainties.
d
The combined standard uncertainty multiplied by the coverage factor (ISO, IEC, OIML, BIPM, 1992; IAEA 1994).
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
38
the national primary standard of air kerma. These
regular TLD calibration measurements should be
carried out at a dose and with an x-ray spectrum
typical of those to which the TLDs will be exposed
during patient-dose measurements. An air kerma
of 10 mGy and an x-ray spectrum generated at a
tube voltage of 80 kV with 3.0 mm Al total filtra-
tion will usually be appropriate for measurements
for conventional radiography.
It will also be necessary to make at least one series
of measurements to establish how the response per
unit of air kerma varies over the entire range of air
kerma, air kerma rates, and x-ray qualities for
which the TLDs are to be used. Ideally, this vari-
ation should be so small that if the response of the
TLDs is determined at one suitable air kerma value,
air kerma rate, and x-ray quality, it could be used
without correction for measurements under any
other conditions, within the required accuracy of
7 %. However, this may not be possible in practice,
and appropriate energy response or other correction
factors may have to be applied to keep the uncertain-
ties within tolerance.
Calibrations of TLDs are usually performed
free-in-air in terms of air kerma. As TLDs applied
in medical x-ray imaging are commonly employed
to measure K
a,e
on patients, calibration on a phan-
tom might be more appropriate. In the latter
case K
a,e
should be derived from K
a,i
obtained
with a reference dosimeter calibrated in terms of
air kerma and employing air kerma based backscat-
ter factors, B according to Eq. (3.15). A set of backs-
catter factors was selected by Petoussi-Henss et al.
(1998) to be used for the calibration of dosimeters on-
phantom. When used for measurement of K
a,e
on
patients, TLDs are usually encapsulated to protect
them against dirt, grease, etc. Consequently, the
TLDs should be calibrated inside the encapsulation
used.
The response of a given TLD can be affected by the
shape and size of the detector and packaging and is
dependent on the direction of the incoming radiation
(McKinlay, 1981; Oberhofer and Scharmann, 1981;
Horowitz, 1984; McKeever et al., 1995), differing
upon exposure free-in-air or in-phantom (Figure 4.1;
Zoetelief et al., 2000). Therefore, the direction of the
radiation during calibration should be the same as
that during the patient dose measurements. For
example, for measurement of the CT axial air
kerma profile for CT using stacks of TLD chips, the
calibration of the TLD chips should be based on cal-
ibration normal to the side of the stack.
0.6
0.65
0.7
0.75
0.8
0.85
0.9
0.95
1
0 15 30 45 60 75 90
Angle of radiation incidence (degree)
R
e
l
a
t
i
v
e

r
e
s
p
o
n
s
e
C80 Free-in-air
C60 Free-in-air
Mo/Mo PMMA-phantom
Mo/Mo Free-in-air
Figure 4.1. Calculated relative response of a 3.2 mm 3.2 mm 0.9 mm LiF TLD-700 detector as a function of orientation of the detector
during irradiation in Mo-filtered 28 kV (Mo anode), 80 kV (C60; HVL
1
2.76 mm Al) and 120 kV (C80; HVL
1
6.31 mm Al) x-ray beams.
0

angle indicates radiation incident perpendicular to the largest surface of the detector. For the 28 kV x rays data are presented for
irradiation free-in-air and for irradiation at a depth of 5 mm in a PMMA phantom. The statistical (type A) standard uncertainties are less
than 1, 2.5, 1, and 1 % for Mo/Mo free-in-air, Mo/Mo in-phantom, C60, and C80, respectively. C60 and C80 are beam qualifications of
Seelentag et al. (1979) (Zoetelief et al., 2000).
MEASUREMENT METHODS
39
4.2 MEASUREMENT METHODS FOR
SPECIFIC DOSIMETRIC QUANTITIES
4.2.1 Dosimeters
For dosimetry in medical x-ray imaging, ioniza-
tion chamber systems and TLD systems are most
commonly used. Suitable ionization chamber dosi-
meters have advantages over TLD systems in that
their accuracy, precision, and energy independence
are better. In addition, ionization chambers can be
read out directly, contrary to TLDs, which have to be
transported to a reading system after irradiation.
Advantages of TLDs are their small size, which
make them suitable for dose measurements on
patients, and their capacity of storing dose informa-
tion over longer periods of time. This makes them
suitable for dosimetry at a distance, for example, at a
central laboratory, TLDs being transported by mail.
Scintillation as well as film dosimeters have also
been used for measurements on patients (Section
4.4), their main characteristics are presented in
this section.
4.2.1.1 Ionization chamber dosimeters
General information on dosimetry using ioniza-
tion chambers can be found elsewhere (ICRU, 1970,
1973; Boag, 1987; IAEA, 1996b). More detailed
information relevant to medical imaging has been
published by Dutreix and Bridier (1985), for
example. Some aspects, specific to medical x-ray
imaging are presented below. Free-in-air air kerma
measurements are best made with suitably designed
ionization chambers of typically between 0.6 and
180 cm
3
volume. The chambers should have air-
equivalent walls so that their energy response in
terms of air kerma is substantially uniform for all
relevant x-ray spectra. The leakage current should
be very small compared with the ionization current
produced by the minimum dose rate to be measured
and the response should not be affected appreciably
by ion recombination at high dose rates. Dosimeters
should be calibrated in a manner traceable to a
national primary standard of air kerma as described
in Section 4.1.
There are special requirements for ionization
chambers used for air-kerma measurements in
mammography: these are a thin entrance wall to
reduce attenuation at low photon energies, and ide-
ally a structure that does not appreciably disturb the
primary radiation field. Thin entrance window
chambers with small volumes generally have a
rather massive construction on the exit side, which
implies that the charge produced in the cavity con-
tains a significant contribution from scattered radi-
ation. Calibration at appropriate mammographic
radiation qualities will, however, overcome the lat-
ter problem. Ionization chambers and their energy
response for mammography beams have been pub-
lished by DeWerd et al. (2002).
4.2.1.2 TLDs
Discussion of general aspects of thermolumines-
cent dosimetry can be found elsewhere (McKinlay,
1981; Oberhofer and Scharmann, 1981; McKeever
et al., 1995; Zoetelief et al., 2000, 2003b). TLDs
have to be read out and annealed carefully to obtain
reliable results. TLDs can be stuck directly and
unobtrusively to the patients skin with very little
interference in patient mobility or comfort. They will
adequately measure the radiation backscattered
from the patient, and they are unlikely to obscure
useful diagnostic information on images made. TLDs
are available in a variety of physical forms and in
different materials. Solid chips are the most conveni-
ent form for application to dosimetry in medical
x-ray imaging.
Lithium fluoride and lithium borate are the most
commonly used phosphors for dosimetry in medical
imaging, combining the desirable features of a relat-
ively flat energy response, low fading characterist-
ics, and reasonable sensitivity. With careful use and
proper calibration they should be capable of measur-
ing doses down to 500 mGy with an expanded uncer-
tainty of <10 % (with a coverage factor of 2). At
100 mGy, the expanded uncertainty is increased to
<25 %. Smaller uncertainties at low doses can be
achieved with the considerably more sensitive cal-
cium fluoride phosphor, but this material is less air-
equivalent and exhibits a marked variationinenergy
response that requires careful calibration at appro-
priate x-ray qualities to achieve accurate measure-
ments. The sensitivity and the accuracy of lithium
fluoride TLDs at doses <50 mGy can be considerably
improved by using deconvolution techniques to sep-
arate out the individual peaks in the glow curve (Bos
et al., 1993, 1994). Alternatively, high-sensitivity
material doped with magnesium, copper, and phos-
phorus (Wu et al., 1984; Fill and Regulla, 1998; Sa ez
et al., 1999) can be used. The main characteristics
(see, e.g., Vij, 1993; McKeever et al., 1995) of four
TLD phosphors are shown in Table 4.5. The energy
dependence of six commonly used chips is shown in
Figure 4.2 (Zoetelief et al., 2000).
4.2.1.3 Scintillation dosimeters
General characteristics of scintillation dosimeters
can be found in Knoll (2000), for example. The
energy dependence of plastic scintillation dosimet-
ers in the energy range used in radiology has been
studied by several authors. Wagner and Pollock
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
40
(1999) found for this type of dosimeter a sensitivity
variation as a function of energy within 5 %
between peak tube voltages of 60 and 125 kV. This
value is confirmed by De Sousa et al. (2000), who
found a sensitivity variation of 10% between tube
voltages of 60 and 140 kV and 25 % between 50 and
140 kV. According to Waite and Fitzgerald (2001) the
variation of the sensitivity of scintillation dosimeters
above tube voltages of 90 kV was similar to that of a
calibrated ionization chamber but deviated at lower
tube voltages values.
The sensitivity of a scintillation dosimeter as a
function of absorbed dose has also been investigated.
Linear dependence of the signal as a function of dose
was observed for absorbed dose up to 60 mGy (De
Sousa et al., 2000) and up to 20 Gy (Wagner and
Pollock, 1999). For doses <0.3 mGy the accuracy
markedly decreases (De Sousa et al., 2000).
4.2.1.4 Film dosimeters
General characteristics of film dosimeters can be
found in Knoll (2000), for example. Two types of film
have been explored for patient dosemetry in inter-
ventional radiology, i.e., radiological film, with a
year-long tradition in this field, and radiochromic
film, which has been introduced more recently.
The radiological film dosimeter is known to have a
very poor tissue-equivalence, because of the high
atomic number of the film emulsion. Two types of
slow KODAK films have been investigated with
respect to their use in interventional radiology. The
X-OMAT-V film, with a long tradition as a radiother-
apy portal dose verification film, shows an increase
in sensitivity by a factor of 20 at peak tube voltage of
70 kV compared with
60
Co gamma rays (1.17 and
1.33 MeV) (Van o et al., 1997). The more recent
EDR2 film does not show a significant variation in
sensitivity between peak tube voltages of 60 and
110 kV (Guibelalde et al., 2003).
Using X-OMAT-V films doses of up to 500 mGy can
be explored and the corresponding linear dose range
is from20 to 200 mGy. For the EDR2 film, doses up to
1000 mGy can be measured and the linear dose range
is from 50 to 500 mGy (Guibelalde et al., 2003). The
EDR2 film is, thus, more suitable for higher-dose
radiological procedures than the X-OMAT-V film.
The tissue-equivalence of the radiochromic film is
considerably better than that of the radiological film,
consequently less variation of its response as a func-
tion of energy is expected. Nevertheless, a sensitivity
variation of up to 5 % between peak tube voltages of
60 and 100 kV has been observed by Giles and
Murphy (2002). For the radiochromic film dosimeter
the dose dependence has been found to be linear up
to an absorbed dose of 10 Gy (Giles and Murphy,
2002). This type of film is thus particularly suited
to evaluate the high doses delivered to the most
heavily irradiated regions, for example, the skin
during some interventional radiological procedures.
4.2.2 Incident air kerma
There are usually several steps required to deter-
mine incident air kerma. The most commonly used
method of determining K
a,i
relies on the measure-
ment of the x-ray tube output, Y(d) (Sections 2.5 and
3.2.1). To derive K
a,i
, Y(d) has to be corrected for the
focal spot-to-surface distance, d
FSD
, and combined
with the recording of the post-exposure P
It
after
examination of a patient or a phantom. The method
is usually suitable only for x-ray machines with a
post-exposure display of P
It
or units with manual
exposure only. Alternatively, K
a,i
can be derived
from a measurement of K
a,e
by applying an appro-
priate backscatter factor [Eq. (3.15); Appendix A].
The incident air kerma, K
a,i
, can be derived from
the air kerma free-in-air at a distance, d, from the
x-ray tube focal spot, K
a
(d)
K
a;i
K
a
d
d
2
d
2
FSD
, 4:2
0.1
1
10
100
10 100 1000
Photon energy (keV)
R
e
l
a
t
i
v
e

T
L

r
e
s
p
o
n
s
e
,

n
o
r
m
a
l
i
s
e
d

t
o

1

M
e
V

p
h
o
t
o
n
s


CaF
2
:Dy
CaSO
4
:Dy
Al
2
O
3
:C
Li
2
B
4
O
7
:Mn,Si
LiF:Mg,Ti
BeO
Figure 4.2. Calculated energy responses relative to air of some
commonly used TL chips. Calculations were performed, assuming
a thickness of 0.9 mm for all detectors (Zoetelief et al., 2000).
MEASUREMENT METHODS
41
where d
FSD
is the focal spot-to-surface distance.
Employing Eq. (2.1) the following relation is
obtained
K
a;i
Y d P
It
d
2
d
2
FSD
: 4:3
The output can be measured free-in-air in the
centre of the x-ray beam at a distance, d, from the
focal spot (d is usually 1 m). Measurements of K
a
(d)
should preferably be made in the manual exposure
mode for the range of exposure conditions met in
clinical practice and the corresponding values of P
It
recorded. In mammography, measurements are
often not made in the centre of the field but closer
to the chest wall edge (EC, 1996b). Because the air
kerma rates can differ appreciably for the same
exposure conditions from one x-ray tube to another,
it is necessary that output measurements are made
on the x-ray tube under study.
For a specific type of radiograph of a patient, for
example, for a posteroanterior (PA) view of the
chest and the exposure conditions (tube voltage,
anode material and angle, filtration) used in practice
for the type of radiograph in question, the post-
exposure P
It
is obtained. The relevant value of Y(d)
is derived from a series of output measurements at
various x-ray tube voltages and filtrations, if neces-
sary by interpolation. The value of K
a,i
for a particu-
lar radiograph of a patient can then be derived
according to Eq. (4.3).
If automatic exposure control (AEC) is used and
the post-exposure P
It
is not indicated, K
a,i
can only
be determined by using a phantom (Section 4.3)
representing the patient. The phantom should pro-
vide the same amount of attenuation as an average
patient and, ideally, should result in the same
x-ray spectrum incident on the AEC device detector.
The air kerma measurement should be made at some
distance from the phantom surface so as to exclude
the contribution from backscattered radiation
and the chamber should not shield the AEC detector.
The incident air kerma K
a,i
can then be calculated
using Eq. (4.2).
4.2.3 Entrance surface air kerma
Measurements of K
a,e
can be made directly on the
patient or a phantom with TLDs. The practical and
theoretical problems related to the use of ionization
chambers for the measurement of K
a,e
are consider-
able. When using phantoms, it may be necessary to
recess a larger dosimeter, for example, an ionization
chamber, in the surface of a phantom so that its
effective centre lies in the plane of the surface and
provides good results (Harrison, 1982; Petoussi-
Henss et al., 1998).
The recommendation in specific radiotherapy
dosimetry protocols for low-energy x rays (e.g.,
IAEA, 1987; IPSM, 1991; Ma et al., 2001) is to derive
K
a,e
from measurement of K
a,i
and application of an
appropriate backscatter factor through Eq. (3.15).
The same approach can be used here. Backscatter
factors for radiology range from 1.25 to 1.60 for
typical x-ray spectra and field sizes used for adults
(Petoussi-Henss et al., 1998). An important excep-
tion is for mammography where the lower energy
spectrum and smaller field size result in backscatter
factors of 1.10. The derivation of appropriate val-
ues for the backscatter factor is discussed in more
detail in Appendix A.
The entrance surface air kerma K
a,e
for a patient
or a phantom for a specific type of radiograph, for
example, for a PA view of the chest or for a medio-
lateral oblique view of the breast, can be obtained as
follows. TLDs should be encapsulatedto protect them
from dirt and exposure to visible light and to facilit-
ate fixing to the surface of a patient or phantom. The
encapsulationmaterial should be thinand should not
contain elements of high atomic numbers, which are
often used in dyes, to avoid significant absorption of
the low-energy x rays sometimes encountered in
radiology, for example, in mammography. For the
relevant radiation quality, the TLDs should be cal-
ibrated inside the encapsulation and at the same
orientation with respect to the x-ray beam as during
medical exposure. The encapsulated TLDs should be
positioned on the patients skin or on top of the
phantom as close as possible to where the centre of
the x-ray beam enters the patient or phantom. Then
the examination should be made in the normal way.
For mammography it has been proposed (EC, 1996b)
that the TLDs be positioned on the upper inner
quadrant of the breast, where the risk of obscuring
clinically important details is less than in the centre
of the beam. When too low a read-out is expected, for
example, for chest radiography, repeated exposures
of the TLDs should be made. Repeated exposures of
the same TLDs should be ideally made for a sample
of sufficient size to represent average patients. The
read-out of the TLDs corrected for background and if
necessary fading and multiplied by the calibration
coefficient results in the value of K
a,e
.
4.2.4 Air kermaarea product
Quantities P
KA
and
_
PP
KA
are most easily measured
using a specially designed ionization chamber,
which is mounted on the collimator housing, where
it affords minimum interference with the medical
imaging task. The chambers are of parallel-plate
design, set perpendicular to the beam axis and are
of sufficient area to encompass the largest beamsize.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
42
The plates are often made to be transparent to vis-
ible light to allow use of a light beam collimator. The
electrometer and display unit are connected to the
chamber by a long cable so that the display can be
positioned for easy access to read and reset. The
response of the P
KA
meter will depend on whether
the chamber is installed on an overtable or under-
table x-ray tube (Section 4.1.2).
The requirements for air kermaarea product
(P
KA
) meters are similar to dosimeters in general,
as described in Section 4.1. This means that
re-calibration should be made at intervals not exceed-
ing 2 years, and whenever the instrument has been
repaired or its performance is suspect (AAPM, 1992).
Re-calibration at one radiation quality, the reference
quality, is often sufficient, as it is unlikely that
the dependence of the response on radiation energy
would change significantly. However, sometimes it
will be important to check comprehensively the
dependence of the response on the dose rate, particu-
larly if the dosimeter is used at widely varying dose
rates. Corrections for ion recombination are usually
insignificant but must be considered when using
very high instantaneous dose rates. Simple constancy
checks and checks of the various functions are recom-
mended before each use of the instrument.
Estimates of the uncertainty in the (in situ) calib-
ration coefficient of a P
KA
meter are 6 %at the 95 %
confidence level (Shrimpton and Wall, 1982; Larsson
et al., 1996) for overtable geometry. For undertable
geometry, the table may reduce the calibration coef-
ficient by 1520 % compared with the overtable
situation (Carlsson and Alm Carlsson, 1990).
Commonly, P
KA
meters consist of PMMA coated
with a conducting material. Carbon is often used for
this purpose as it is approximately air-equivalent,
however, to achieve light transparency other con-
ducting materials are used. These materials con-
tain high atomic number elements resulting in a
relatively strong energy dependence compared with
ionization chambers coated with graphite (Larsson
et al., 1996; BednarekandRudin, 2000). For example,
the addition of additional copper filtration may
affect the calibration coefficient by several per cent.
Experience with ageing of air kermaarea product
meters, for example, when installed at interven-
tional radiology systems and submitted to very
large doses during several years is still limited. Sim-
ple constancy checks of the instrument should be
able to reveal ageing.
Some P
KA
meters are capable of distinguishing
between values accumulated during radiography
and fluoroscopy. If this distinction is made purely
on the basis of a threshold in
_
PP
KA
(difference in
tube current exposure time product rate between
fluoroscopy and radiography), errors can occur. For
example, when large beam areas are used during
fluoroscopy of thick or dense areas of the patient,
_
PP
KA
approaches values seen during radiography.
Developments are in progress where the field size,
d
FSD
and K
a,i
are also registered. Modern x-ray
equipment may have built-in P
KA
meters or will dis-
play the value of P
KA
after an exposure by internal
calculation using the exposure factors (tube voltage
and filtration, P
It
) and the position of the beam lim-
iting collimators. The value of P
KA
does not contain
information on the projection.
The assessment of organ doses from P
KA
measure-
ments for procedures, involving both fluoroscopy
and radiography, is quite complex. This is illustrated
for colon examinations where the value of P
KA
has to
be related to individual radiographs and to interme-
diate fluoroscopy. Values of
_
PP
KA
and cumulative P
KA
are shown in Figure 4.3. The standard colon exam-
ination was biphasic (Geleijns et al., 1997). First,
Figure 4.3. Air kermaarea product rate,
_
PP
KA
(top panel) and
cumulative air kermaarea product, P
KA
(bottom panel) as a
function of examination time (sample rate 0.1 s) measured during
a colon examination of one patient (Reproduced from Geleijns
et al., 1997, with permission from Radiological Society of North
America). The P
KA
values have to be related to images as presen-
ted in Table 4.4 and to fluoroscopy. The data in the present figure
correspond to Patient 4 in Figure 4.4.
MEASUREMENT METHODS
43
after insertion of a rectal cannula, a barium suspen-
sion of low density and high viscosity was admin-
istered and a single contrast study was performed.
Second, after partial evacuation, air was insufflated
for the double contrast study of the rectum and the
colon. From this information it has to be determined
whether the increase in
_
PP
KA
belonged to a continu-
ous, prolonged signal, which is indicative of fluoro-
scopy, or to a short pulse, which is indicative of
radiography (imaging). If radiography was recog-
nized, the P
KA
value for the radiograph was calcu-
lated from all P
KA
samples in the pulse, which
included the P
KA
sample acquired at the start of
the pulse (for which the value might be low).
Although a standard protocol for colon examination
was implemented (Table 4.6), deviations from the
protocol were common and had to be taken into
account during analysis. The P
KA
for each radio-
graphic projection (image) and each fluoroscopy
interval can be estimated only after careful assign-
ment of the corresponding projection (Table 4.6.) to
each peak in the
_
PP
KA
signal (i.e., to each radiograph).
The distribution of P
KA
over the various projections
is shown in Table 4.6. For fluoroscopy, it is assumed
that the exposure conditions can be mimicked by the
subsequent radiograph, except for differences in
radiation quality (i.e., tube voltage). Consecutive
P
KA
for fluoroscopy and radiographs (images) during
colon examination of 13 patients are shown in
Figure 4.4. The figure illustrates the differences
occurring in practice even when a standard protocol
is implemented.
4.2.5 CT air-kerma index and CT air-kerma
index in the standard CT head and
body dosimetry phantoms
Following the convention established in Section
3.2, C
K
refers to measurements free-in-air and C
K,p
to measurements in-phantom.
4.2.5.1 Pencil ionization chamber dosimeter
For dosimetry in CT, pencil ionization chambers
are available with a diameter of 10 mm, a sensitive
length of 100 mm, and an active volume of 3 cm
3
.
An essential feature of these dosimeters is that their
response is designed to be uniform along the entire
length of the sensitive volume. They are usually cal-
ibrated in terms of air kerma by exposing the entire
sensitive volume to a uniform x-ray field. Thus they
indicate P
KL
/L when exposed to a CT slice or mul-
tiple slices free-in-air. Because the C
K
of Eq. (3.22)
rapidly falls to zero away from the position of the CT
slice, the pencil ionization chamber will also meas-
ure C
K
to a very good approximation. The relation-
ship between C
K
, and P
KL
/L, the single slice CT axial
air kerma profile K
a
(z), the chamber length L and
the nominal slice thickness T is shown in Figure 4.5.
By definition, the area under the single slice dose
profile (the dotted line) equals the areas of either
rectangle. For measurement inside a phantom, C
K,p
is obtained in a similar way.
For a measurement of C
K
, the pencil ionization
chamber should be fixed so that it extends beyond
the table end. The chamber should be mounted in
0
5
10
15
20
25
30
1 2 3 4 5 6 7 8 9 10 11 12 13
A
i
r

k
e
r
m
a

a
r
e
a

p
r
o
d
u
c
t
;

P
K
A

/

G
y
.
c
m
2
Patient number
Figure 4.4. Air kermaarea products, P
KA
values for fluoroscopy (white) and images (black) during colon examination of 13 patients
(Geleijns, 2000, private communication). The data for the 13 patients have been ordered in accordance with the total P
KA
for the
examination. Reproduced with permission from Dr. J. Geleijns.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
44
such a way that its length axis corresponds to the
axis of rotation of the gantry and that its centre
corresponds to the centre of the slice, or the centre
of the slices for a multi-slice scanner (Figure 4.6).
The measurement with the ionization chamber
should be made for a single rotation. When the rota-
tion differs from360

this should be noted. The value


of C
K
is obtained from the ionization chamber read-
ing as the air kerma multiplied by the length of
the chamber, L, and divided by the nominal slice
thickness, or the sum of the nominal slice thick-
nesses in the case of multi-slice scanners.
For a measurement of C
K,PMMA
, the CT dosimetry
phantoms (Figure 4.7) should be used. The phantom
should be fixed on the table in such a way that its
longitudinal axis coincides with the axis of rotation
of the gantry and that its centre corresponds to the
centre of the slice, or the centre of the slices for a
multi-slice scanner. The pencil ionization chamber
should be inserted in the phantomin such a way that
its centre corresponds to the longitudinal central
plane of the phantom for each of the five measure-
ment positions, i.e., one at the centre and four at
1 cm below the surface (Figure 4.7). Unused holes
should be plugged.
For radiotherapy, similar problems arise for meas-
urements inside phantoms with ionization chambers
calibrated in terms of air kerma free-in-air as for
dosimetry for x rays used in medical imaging.
According to the IAEA (1996b) an ionization cham-
ber calibrated for the x-ray quality considered can be
used to derive the absorbed dose to water in a water
phantom by using the relation
D
w
M
u
N
K
mm
en
r

w;a
k
a;w
, 4:4
where M
u
is the dosimeter reading corrected for
influence quantities such as temperature and pres-
sure, N
K
the air kerma calibration coefficient for the
x-ray quality considered, mm
en
=r
w;a
the ratio of the
mean mass energy absorption coefficients of water to
that of air averaged over the energy fluence distri-
bution of the photons at the point of measurement in
gantry
gantry
pencil ionization chamber
axis of rotation
table
Figure 4.6. Schematic diagram of the arrangement for
measurement of C
K
.
a
b d
A B
c
Figure 4.7. Standard dosimetric phantoms for CT of the body
(A) and the head (B) [ICRU Report 48 (1992c)]. The PMMA
phantoms representing the body and the head have diameters of
32 cm (a) and 16 cm (c) and minimum lengths of 14 cm (b, d). They
contain cavities at the centre and the periphery to allow the
insertion of pencil ionization chambers.
0 - /2 L
DISTANCE TO CENTER OF SLICE
C
K
T
K
a
(z)
P L
KL
/
, z
Figure 4.5. A single slice CT axial air kerma, K
a
(z) profile with the
corresponding values of C
K
, P
KL
/L, L, and T, where L and T are
the chamber length and the nominal slice thickness, respectively.
By definition, the area under the single slice dose profile (the
dotted line) is equal to the area of either rectangle.
MEASUREMENT METHODS
45
the water phantom, and the overall correction k
a,w
comprises the following components: k
E,u
the correc-
tion factor for the energy and angular dependence of
the response of the ionization chamber, p
d
the dis-
placement correction factor which accounts for the
effect of the displacement of water by the air volume
of the shape of the ionization chamber, k
st
the cor-
rection factor for the influence of the stem of the
ionization chamber, k
sl
the correction factor, which
accounts for the effect of the protective sleeve needed
if a non-watertight ionization chamber is inserted
into a water phantom, k
b
is the correction factor for
yet unknown effects, the value of which is taken to
be unity.
Both the IEC (1999) and the EC (2000) recommen-
ded to measure the absorbed dose to air inside the
standard CT head and body phantoms. For a meas-
urement inside an air cavity within a PMMA phan-
tom the absorbed dose to air varies over its volume.
The exact nature of this variation and hence also
of the absorbed dose to air inside the cavity depends
on the range of the electrons produced in photon
interactions. For three x-ray qualities (IEC, 2004),
this illustrated in Figure 4.8 for the absorbed dose
to air in air at a PMMA/air interface. When the
commonly recommended detector, i.e., an ionization
chamber has a wall thickness sufficient to absorb the
most energetic electrons produced inside the PMMA,
air kerma can be measured inside the PMMA phan-
tom. This is schematically shown in Figure 4.9.
When measurements are made inside a PMMA
phantom, Eq. (4.5) applies for the air kerma
measured in the phantom, K
a,PMMA
is
K
a;PMMA
M
u
N
K
k
a;PMMA
, 4:5
where k
a,PMMA
is in first approximation equivalent
to k
a,w
. The overall correction factor for a given ion-
ization chamber, k
a,PMMA
, accounts for the differ-
ences between measurements free-in-air and at
depth in a PMMA phantom. The numerical values
of k
a,PMMA
and k
a,w
for an NE2571 (Nuclear Enter-
prises, Reading, UK) ionization chamber are expec-
ted to differ by <0.5 %as the displacement correction
factor for stemless chambers is decreasing by 0.5 %
between densities of 1.00 and 1.17 g cm
3
for x rays
with HVL
1
in excess of 3 mm Al (Ma and Nahum,
1995). The dependence of k
a,w
of an NE2571 ioniza-
tion chamber on field size and depth for 80 kV x rays
at a HVL
1
of 2.4 mm Al is small, i.e., 1 %
(Seuntjens and Verhaegen, 1996). Numerical values
of k
a,w
are almost only available for the NE2571
ionization chambers. A k
a,w
value for the NE2571
ionization chamber of 1.01 has been published by
the NCS (1997) for tube voltages between 80 and
120 kV and a HVL
1
of 2.8 mm Al. Values are lacking
for the pencil ionization chambers used for dosi-
metry in CT. It is, however, expected that k
a,PMMA
are only a few per cent different from unity as the
diameters of the pencil chambers and the NE2571
ionization chamber are not too much different.
The C
K,PMMA
value follows from the ionization
chamber reading multiplied by the chamber calib-
ration coefficient for in-phantom measurements of
Figure 4.9. Schematic diagram of the variation of the absorbed
dose to air and of the air kerma and the substance m kerma at
an interface of material m and air with effective atomic numbers
Z
eff,m
< Z
eff,air
. Courtesy: H.M. Kramer (2002), private
communication. The right for publication was obtained from the
Physikalisch-Technische Bundesanstalt.
Figure 4.8. Illustration of the absorbed dose to air (left ordinate,
solid lines) and kerma to material m (right ordinate, broken lines,
m air, PMMA) in the vicinity of a PMMA/air interface. All
quantities are normalized to K
air
1. Black, red, and blue
curves refer to radiation qualities at tube voltage and HVL
1
of
70 k and 2.5 mm Al, 100 kV and 3.97 mm Al, and 150 kV and
6.57 mm Al, respectively. The kerma to material m (m air,
PMMA) has a discontinuity at the interface. Courtesy: H.M.
Kramer (2005), private communication. The right for publication
was obtained from the Physikalisch-Technische Bundesanstalt.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
46
air kerma and the length of the chamber, L, divided
by the nominal slice thickness, or the sum of the
nominal slice thicknesses in the case of a multi-
slice scanner. Note that for multi-slice CT, the use
of a 100 mm long ionization chamber may in some
cases be less appropriate than for single slice CT.
This is because of the increase in irradiated com-
bined slice thickness associated with the multi-slice
technique. When the rotation differs from 360

this
should be noted. Special attention should be given to
the orientation of the peripheral measurement posi-
tions. Thus the C
K,PMMA,100,c
can be obtained from
the measurement at the central position and
C
K,PMMA,100,p
as the average of measurements at
four peripheral positions. By using Eqs (3.263.28),
the CT air kermalength product, P
KL,CT
for a com-
plete conventional (serial CT) examination can be
derived from C
K,PMMA,100,c
and C
K,PMMA,100,p
. In the
case of helical scanning, P
KL,CT
can be derived from
C
K,PMMA,100,c
and C
K,PMMA,100,p
by using Eqs (3.26),
(3.27), and (3.29).
4.2.5.2 TLDs
TLDs placed in a row within a suitable plastic
holder can be used to measure the CT axial air
kerma profile free-in-air. The TLDs should be con-
tiguous in the region of the primary beam, but away
from this region, they can be separated by spacers. It
is noted that TLDs can be arranged in a stack or
placed side by side. The latter arrangement is applic-
able for measurement of C
K
but does not give a good
measurement of the air kerma profile. In principle
TLDs can also be used for measurements in phan-
toms, but if several measurement positions are used
then this will require a large number of dosimeters.
Therefore, ionization chambers are preferred for in-
phantom measurements. A typical free-in-air single
slice dose profile measured with lithium fluoride or
lithium borate TLD chips, 0.8 mm thick and stacked
contiguously in a hollow cylindrical plastic capsule,
is shown in Figure 4.10A. For such a stack of n TLDs,
the C
K
is approximated by
C
K

1
T
Z
1
1
K
a
z dz
1
T
S
n
K
aj
h, 4:6
where K
aj
is the air kerma to the jth TLDfroma total
of n TLDs and h is the extension (0.8 mm in this
example) of each TLD in the axial direction, and T is
the nominal slice thickness.
The value of C
K
, thus derived, is shown in
Figure 4.10B; for this particular scanner and slice
thickness, is 8 % higher than the peak of the dose
profile.
For the measurement of C
K
, the stack of TLDs
should be fixed so that it extends beyond the table
end (similar to the fixation of the pencil ionization
chamber shown in Figure 4.6). The stack of TLDs
should be mounted so that its length axis coincides
with the axis of rotation of the gantry and that its
centre corresponds to the centre of the slice, or the
centre of the slices for a multi-slice scanner. The
measurement with the stack of TLDs should be
Figure 4.10. Typical free-in-air CT axial air kerma profile K
a
(z) for a nominal slice thickness of 4 mm measured by TLDs (A) and the
resulting C
K
(B) (after Shrimpton et al., 1991; Survey of CT Practices in the UK, Part 2: Dosimetric Aspects, NRPB-R249; HMSO, London,
UK).
MEASUREMENT METHODS
47
made for a single rotation. When the rotation differs
from 360

, this should be clearly stated.


4.3 FEATURES OF MEASUREMENTS ON
PATIENTS AND MEASUREMENTS WITH
PHYSICAL PHANTOMS
Measurements of K
a,i
, K
a,e
, and P
KA
can be made
in the presence of patients or by using phantoms.
Measurements with patients have the advantage
that no specific phantoms have to be designed and
constructed. Patients vary in size, which will influ-
ence the measurement result. For mammography,
the tissue composition of the female breast, i.e., the
fraction of glandular tissue, is of importance. To
obtain a representative average value of, for
example, K
a,i
, measurements have to be made for a
series of patients. Such measurements provide
insight into the dose variation among individual
patients and of actual clinical practice (Gentry and
DeWerd, 1996). To limit the number of patients, a
selection can be made on body mass (EC, 1996a) or in
the case of mammography, on compressed breast
thickness (EC, 1996a,b). For organ dose assessment,
it will be necessary to obtain information on the
gender of the patients involved, as the dose-
conversion coefficients can be gender-specific. Also
the age of the patients involved should be recorded.
An anthropomorphic (or body) phantom has the
shape of a human body or part of it (ICRU, 1992c).
It is constructed of various tissue substitutes
simulating the human body with respect to size,
shape, mass density, and radiation attenuation and
scattering. The use of anthropomorphic phantoms
for assessment of organ doses is discussed in
Section 5.1.
Phantoms are particularly useful for surveys of
patient doses and for quality control. For example,
the US Center for Devices and Radiological Health
(CDRH) has designed phantoms for specific exam-
inations, i.e., the CDRH chest phantom (Conway
et al., 1984), and the CDRH abdomen and lumbo-
sacral spine phantom (Conway et al., 1990). Both
phantoms are made out of PMMA and aluminium.
Havukainen and Pirinen (1993) from the Finnish
Radiation and Nuclear Safety Authority (STUK) fol-
lowed a similar approach. They constructed PMMA-
Al phantoms for examinations of abdomen PA, chest
PA, lumbar spine PA, skull LAT and thoracic spine
anteroposterior (AP). The CDRH and STUK phan-
toms are designed to provide characteristics similar
to that of body parts of patients, with respect to both
attenuation and beam hardening. However, they do
not allow for dose measurements at depth in the
phantom.
Standard dosimetric phantoms have simple repro-
ducible geometry and are used to compare measure-
ments under defined conditions of irradiation.
Standard phantoms as proposed by the ICRU
(1992c), but made out of PMMA, are used for dosi-
metry for mammography and for CT. The standard
mammography phantom described by the ICRU has
a thickness from 2 to 8 cm. However, often only one
thickness is recommended, for example, 4.5 cm (EC,
1996b). The cross-sectional dimensions recommen-
ded by the EC (1996b) are somewhat larger than
that described by the ICRU, i.e., either semicircular
with a radius of at least 100 mm or rectangular with
an area of at least 150 mm 100 mm.
It should be borne in mind that the phantoms
at best represent (parts of) average-sized adult
patients, and, therefore, do not allow accurate dose
assessments for individuals of different sizes or body
composition. However, the use of a phantom that
mimics the attenuation and beam hardening in a
specific part of a standard patient correctly has the
advantage that the number of measurements
required to derive the incident air kerma is consid-
erably smaller than that for measurements with
patients.
4.4 SKIN DOSE DETERMINATION
A growing number of publications report on
deterministic effects in the skin of patients who
have undergone interventional radiological proced-
ures (Sovik et al., 1996; Van o et al., 1998, 2001a;
ICRP, 2000; Geleijns and Wondergem, 2005; Van
Sonnenberg et al., 2005; Wambersie et al., 2005).
Measurement of the maximum skin dose, also
referred to as peak skin dose (Section 1.4), is, there-
fore, increasingly important. The main issue is that
the location of the most heavily irradiated regions of
the skin is not known in advance. In principle three
methods can be used for skin dose measurements.
4.4.1 Direct measurement of the
maximum skin dose
The most straightforward method for skin dose
determination is to put TLDs or other detectors on
the patients skin. However, the choice of the meth-
ods and dosimeters are not yet clearly established. In
this section an overview is given of the measurement
of the maximum skin dose D
skin,local
during radiolo-
gical examinations. Although most of the measure-
ments were actually performed with TLDs,
scintillation dosimeters and film dosimeters have
also been used. Semiconductor dosimeters [diodes
or metal oxide semiconductor field effect transistors
(MOSFETs)] have been used only for phantom
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
48
measurements. These latter dosimeters have not
been used on patients because the filters, needed to
compensate for the poor tissue-equivalence, appear
on the radiological image. Therefore, they are not
further discussed here.
4.4.1.1 Skin dose measurements on patients
with TLDs
The basic characteristics of TL dosimeters are
presented in Section 4.2.2. Numerous skin dose
measurements were made with the standard phos-
phor LiF:Mg,Ti in diagnostic and interventional
radiology (Marshall et al., 1995; Ng et al., 1998;
Kemerink et al., 2001; Papageorgiou et al., 2001;
Van o et al., 2001a). The wide application of this
phosphor in personal dosimetry and in in vivo dosi-
metry in radiotherapy (Van Dam and Marinello,
1994) led to its application in radiology. It is com-
mercially available, its characteristics are well esta-
blished, and its energy dependence is acceptable.
Lithium borate, with weaker energy dependence
in the radiological range (Table 4.5 and Figure 4.2),
has also been used on patients (Maccia et al., 1988;
Gregan et al., 1998).
Table 4.5. Main characteristics of four TLD phosphors
a
(see, e.g., Vij, 1993; McKeever et al., 1995).
Phosphor Li
2
B
4
O
7
LiF LiF CaF
2
Dopant Mn Mg, Ti Mg, Cu, P Dy
Relative energy response (30 keV/1 MeV) 0.95 1.35 1.3 14.5
Fading
b
5 % month
1
<5 % y
1
<5 % y
1
1 % d
1
Relative sensitivity
c
to 1 MeV photons 0.5 1 30 15
Detection threshold (mGy) 100 50 5 1
a
All four phosphors are commercially available as chips (dimensions: 3.2 mm 3.2 mm 0.9 mm).
b
Fading % loss at 20

C per year (y), month or day (d).
c
Read-out per unit air kerma compared with that of LiF: Mg, Ti.
Figure 4.11. Three-dimensional representation of the patient skin dose distribution (mGy) as measured with a grid of TLDs for a coronary
angiography in a biplane setting. Results from a Belgian multi-center study on skin doses of patients treated with prolonged cardiological
interventions. Study sponsored by the Federal Agency of Nuclear Control in Belgium. With permission from E. Bogaert, K. Bacher,
H. Thierens.
MEASUREMENT METHODS
49
Two studies in interventional radiology deserve
special attention, because they illustrate the pos-
sibilities of TLD for tracing the location of the
maximum skin dose D
skin,local
for complicated pro-
cedures. For 137 patients, undergoing various inter-
ventional procedures, Waite and Fitzgerald (2001)
placed LiF chips in arrays of 35 chips on the patients
skin and determined D
skin,local
by interpolation. For
the same purpose, Geise et al. (1997) explored the
suitability of magnesium borate TLD arrays as
monitors in potential high-dose x-ray areas. After
irradiation, the signal is obtained through laser
stimulation. The dosimeters are arranged in radio-
lucent arrays covering a surface of 30 cm 30 cm,
with dots of TLD material 2 mm in diameter
positioned at 3 mm intervals in rows, spaced 3 mm
apart.
Results from a Belgian multi-center study on
assessment of local skin dose of patients treated
with prolonged cardiological interventions are shown
in Figure 4.11. Arrays of TLDs that are wrapped
around the patient (with spacing larger than 3 mm
for practical reasons) allow visualizing of the three-
dimensional distribution of the patient skin dose.
4.4.1.2 Skin dose measurements on patients
with scintillation dosimeters
The basic characteristics of scintillation dosimet-
ers are presented in Section 4.2.3. One of the poten-
tial advantages of scintillation dosimeters is real-
time information. Although one would expect wide-
spread clinical applications, only a limited number of
studies explored the possibilities of scintillation dosi-
meters in interventional radiology. Wagner and
Pollock (1999) used scintillation dosimetry for mis-
cellaneous high-dose fluoroscopically guided proced-
ures and Hwang et al. (1998) for cardiac
interventional radiology.
4.4.1.3 Skin dose measurements on patients
with film dosimeters
The basic characteristics of film dosimeters are
presented in Section 4.2.4. For complicated proced-
ures, where the position of the peak skin dose is hard
to predict, film dosimeters offer unique possibilities.
When wrapped around the patient, the film provides
information on the dose distribution at the level of
the skin. The energy of the x-ray beam should not
Table 4.6. Example of a standard protocol for biphasic colon examination (Reproduced from Geleijns et al., 1997, with
permission from Radiological Society of North America).
Projection View
a
Nominal tube
voltage/kV
HVL
1
/mm Al d
FID
b
/cm d
FSD
b
/cm Field size on
image cm cm
P
KA
c
/Gy cm
2
Full column, single contrast examination
Rectum and Sigmoid AP 125 7.5 126 97 24 30 0.37
Rectum and Sigmoid LPO 133 7.8 126 93 24 30 0.74
Rectum and Sigmoid Left decubitus LLAT 141 8.0 126 77 24 30 1.91
Rectum and Sigmoid AP (4 radiographs)
d
133 7.8 126 97 12 15 0.31
Splenic Flexure RPO 133 7.8 126 91 24 30 0.55
Hepatic Flexure LPO 133 7.8 126 94 24 30 0.84
Cecum AP (2 radiographs) 125 7.5 126 98 24 15 0.71
Colon Supine AP (survey) 125 7.5 151 122 35 43 1.04
Double contrast examination
Rectum and Sigmoid PA 96 6.1 126 93 24 30 0.45
Rectum and Sigmoid LLAT 109 6.6 126 77 24 30 2.51
Rectum and Sigmoid LPO 102 6.3 126 93 24 30 0.74
Rectum and Sigmoid Caudal-cranial inclination 102 6.3 126 79 24 30 0.92
Rectum and Sigmoid AP 96 6.1 126 97 24 30 0.61
Rectum and Sigmoid RLAT 109 6.6 126 70 24 30 2.95
Splenic Flexure RPO 102 6.3 126 91 24 30 0.83
Hepatic Flexure LPO 102 6.3 126 94 24 30 0.70
Cecum AP (2 radiographs) 96 6.1 126 98 24 15 0.65
Colon Supine AP (survey) 96 6.1 151 122 35 43 1.11
Colon Prone PA (survey) 96 6.1 151 122 35 43 0.73
Biphasic examination, i.e., single and double contrast 20.9
Double contrast views only 12.8
a
AP is anteroposterior, LLAT left lateral, LPO left posterior oblique, PA posteroanterior, RLAT right lateral, RPO right posterior
oblique.
b
d
FID
is focal spot-to-image receptor distance, d
FSD
focal spot-to-surface distance.
c
The air kermaarea products, P
KA
were measured on an overtable x-ray tube, 0.1 mm Cu filtration, fluoroscopy is performed at 110 kV.
d
Air kermaarea product, P
KA
is given for one single exposure.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
50
vary too much over the area of the detector and
during the procedure. These requirements are
more important for radiological film than for the
radiochromic film, as discussed previously (Section
4.2.1.4). The possibilities of the film for measure-
ments on patients were illustrated for complicated
cardiac interventional radiology procedures
(Guibelalde et al., 2003).
Figure 4.12 shows an illustration of the use of slow
films to measure patient skin dose during cardiac
interventional radiology. The use of additional ther-
moluminescent or optically stimulated luminescent
dosimeters at the positions of the film where the
highest doses are expected allows verification the
calibration of the film.
4.4.2 Derivation of the skin dose from the
air kermaarea product P
KA
Skin dose can, under certain conditions, be
derived from the air kermaarea product
P
KA
. There are basically two methods to derive the
skin dose from the air kermaarea product. The first
method determines both quantities independently
on a significant number of patients and derives, by
pooling the data, conversion coefficients for calculat-
ing the skin dose. Waite and Fitzgerald (2001) cor-
related air kermaarea product P
KA
and skin dose
measurements (using TLD arrays) on a population
of patients undergoing percutaneous transluminal
coronary angioplasty (PTCA) and neurological pro-
cedures. The correlation between skin dose and air
kermaarea product was found satisfactory for the
neurological patients but was poor for the PTCA
patient group. Van de Putte et al. (2000) found also
a poor correlation for a population of patients under-
going cardiac procedures. As pointed out by the
authors, this can be explained by the fact that the
cardiac radiological procedures implied more
patient-to-patient differences in geometry.
Relying purely on geometrical factors
(Figure 4.13), the skin dose can be derived from the
air kermaarea product, provided the field area at
the level of the patients skin is known assuming
that conditions of scattered radiation are similar.
This method is straightforward when the collima-
tion, as well as the focus-to-skin distance, is kept
constant throughout the full duration of the exam-
ination. This is valid for simple procedures, but not
for complicated examinations as used in interven-
tional radiology for which multiple angulations,
field sizes, and focus-to-patient distances are often
applied. Although an inherent loss of accuracy can-
not be avoided, useful information can be obtained
by assigning to each type of procedure a nominal
geometry and, on the basis of this geometry, to
derive the skin dose from the air kermaarea prod-
uct readout. For different interventional radiological
procedures McParland (1998) analysed the uncer-
tainties in the skin dose values, caused by realistic
deviations from the nominal geometry, and showed
that the estimated dose values were at least within
40 % and generally within 30 % of those actually
received.
Figure 4.12. Illustration of the use of slow films to measure
patient skin dose. Top: Practical set-up to measure skin dose
distributions. Thermoluminescent or optically stimulated lumin-
escence dosimeters are attached to an envelope that contains a
slow film (KODAK EDR2), in the positions where the highest
doses were expected. The use of a double set of dosimeters was
used to verify the response of both technologies for this kind of
measurements. Dosimeters allowed verifying the calibration of
the film. Bottom: Example of slow film used to measure the skin
dose distribution in a patient during percutaneous transluminal
coronary angioplasty (PTCA). Note the wide distribution of the
radiation fields on the skin owing to the different orientations of
the x-ray beam. Collimation and the use of wedge filters are
clearly visible. Courtesy: E. Van o.
MEASUREMENT METHODS
51
Another approach consists of real-time automatic
data acquisition of the fluoroscopy conditions: time
and geometric parameters. Kicken et al. (1999)
recorded continuously the field area at a reference
distance of 1 m. The field area at the level of the
patients was derived from this value using the
focus-skin distance, obtained electronically from
the focus-to-image receptor distance. The method
was applied on patients undergoing arteriography
of the lower limbs.
4.4.3 Derivation of the skin dose directly
from the radiological parameters of the
exposure
The skin dose was also derived by some authors
relying on the radiological parameters of the
Figure 4.13. Derivation of skin dose fromair kermaarea product and geometrical factors. Top left: The transmission chamber is calibrated
against a reference ionization chamber, which has been calibrated in air kerma free-in-air. The reference chamber is positioned, on the
beam axis, at a reference distance d
0
of 60 cm from the tube focal spot. This point is chosen 20 cm above the position of the patient couch
and corresponds approximately to the point of entrance of the beam axis at the patient skin Beam dimensions of 10 cm 10 cm at the
reference distance d
0
are selected (checked by film). The signal of the transmission chamber is expressed in terms of air kermaarea-
product in Gy m
2
(or convenient multiples). Top right: For distances differing from the reference distance (e.g., d
0
d
x
), the beam area is
increased by [(d
0
d
x
)/d
0
]
2
and the dose on the beam axis is decreased by the same factor. The air kermaarea product thus remains
unchanged. Bottom: The irradiation geometry is modified compared with that in the top: the tube is located under the couch and the
backscatter fromthe patient is introduced. The transmission chamber is calibrated against a reference ionization chamber, which has been
calibrated in absorbed dose to water (tissue), under patient (backscatter) conditions. As indicated above, the air kermaarea product
remains unchanged when modifying the distance between the tube focus and the entrance skin of the patient (for the indicated range).
Courtesy: A. Wambersie.
Figure 4.14. Photograph showing the ionization chamber attached
to the anthropomorphic head phantom used to measure entrance
skin dose [from Gkanatsios et al. (1997), with permission].
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
52
exposure (Section 1.3.4.1). Such an approach needs
phantom calibration before clinical use. Gkanatsios
et al. (1997) introduced a commercial on-line patient
exposure meter that was installed in each plane
of a biplane neuroradiological imaging system
(Figure 4.14). A device, introduced by the authors,
can compute skin dose on the basis of selected tech-
nical parameters such as peak tube voltage, tube
current, and focus-to-patient distance.
Simulations of radiological procedures were also
performed on the phantom with an ionization cham-
ber to measure the systems accuracy by comparing
the computed and measured dose. The deviation
between the two values did not exceed 5 % for a
group of neuroradiological patients.
For similar applications, ODea et al. (1999)
applied a method also based on an automated dosi-
metry system. The standard deviation obtained by
combining the different uncertainties was estimated
to be 10 %.
If fully digital systems are available then a
more automated approach becomes possible,
because the Digital Imaging and Communications
in Medicine (DICOM) header information con-
tains tags for different data with respect to the
particular image. Software tools could extract the
necessary exposure details automatically. Such a
type of system was developed at the Complutense
University of Madrid (Van o et al., 2002).
Figure 4.15 shows an example of the local skin
doses due to different irradiation fields during a
cardiac interventional procedure obtained from
the information contained in the DICOM header
of a series of images. The results are compared
with skin doses derived from the use of a slow
film combined with TLDs.
Total skin dose MAG30712; 08/01/03
-40
-20
0
20
40
-90 -40 10 60
right and left of the patient
c
r
a
n
a
n
d
c
a
u
d
o
f
t
h
e
p
a
t
i
e
n
t
TLD measured in this area (not peak) 2.4 Gy
Calculated
from DICOM
2.7 Gy
Figure 4.15. Peak skin dose assessment using data from the DICOM header compared with results on skin dose using slow film and TLDs.
Top right: A slow film is positioned on the back of the patient to measure the skin dose distribution during an interventional cardiology
procedure guided by fluoroscopy. Top left: The details of the duly processed slow film (with the grey scale inverted) are shown. Note that in
this case, the film is clearly saturated and some TLDs positioned on the film to measure skin dose and to verify the calibration (dose versus
film density) give a dose value of 2.4 Gy. This is not necessarily the peak skin dose if the TLD was not at the position where the maximum
dose is imparted. Bottom left: A diagram with the doses in the different radiation fields obtained from the information contained in the
DICOM header of the series of images is shown. In some cases, this calculation allows to estimate with a better approach the peak skin
dose. In the reported case, the estimation from DICOM header gives a peak skin dose of 2.7 Gy (shown top right). Reproduced with
permission from E. Van o.
MEASUREMENT METHODS
53
Journal of the ICRU Vol 5 No 2 (2005) Report 74 doi:10.1093/jicru/ndi027
Oxford University Press
5 METHODS FOR DETERMINING ORGAN AND
TISSUE DOSES
For the assessment of the dose to patients to be
indicative of the stochastic radiation detriment,
estimates of D
T
for radiosensitive organs and tissues
in the body are required. It is possible to assess D
T
for certain compact and superficial organs such
as the thyroid or testes from measurements with
TLDs at appropriate locations on the patients skin,
although corrections may be necessary to account
for the amount of attenuation through the organ or
through the overlying tissue. In interventional radi-
ology, it will be useful to check D
T,local
(or D
skin,local
)
for the skin in the primary x-ray beam in relation to
the threshold for induction of deterministic effects.
TLDs can be used to assess the local skin dose,
although predicting the location of the maximum
dose is difficult.
More generally, D
T
or D
T,local
can be measured
using appropriate physical models of the human
body that can be loaded with dosimeters and
exposed under conditions typical of particular exam-
inations or procedures. Alternatively, computer
programs are available that simulate radiation
transport using a computational model of the
patient. They can be used to calculate D
T
with
much greater speed and flexibility than is possible
by measurement. Extensive tabulations and data
files on computer disk are available containing the
results of these calculations in the form of coeffi-
cients for converting the more readily measurable
normalization quantities (Section 3.4) into organ
doses for specific x-ray exposure conditions in
medical imaging.
Because these assessments are usually made
for particular practices or techniques that are
applied to a large number of patients, there is not
the same need for individually tailored patient-dose
estimates as occurs in radiotherapy. As a conse-
quence, models for standard patients are frequently
used for dosimetry in medical x-ray imaging.
Both physical and computational phantoms have
been developed to represent adult and paediatric
patients. When individual patient dose determina-
tions are required, e.g., for accidental overexposures,
the calculations based on standard phantoms are
useful as a first estimate. For an improved estimate,
the particular anatomical details of the patient
and the specific exposure conditions need to be
considered.
5.1 DOSE MEASUREMENTS IN PHYSICAL
PHANTOMS
Physical dosimetry phantoms are designed to
simulate the way in which a patient, or part of a
patient, will absorb and scatter ionizing radiation.
Measurements using the phantoms can then be rep-
resentative of the dose distributions occurring in
clinical practice. Phantoms can be of varying degrees
of complexity and anatomical accuracy depending on
the purpose of the dose determinations (Section 4.3).
If the doses received by a number of radiosensitive
organs within the patient are to be measured, an
anthropomorphic phantom that accurately models
human anatomy is needed. Such phantoms are com-
mercially available but it is important to be sure that
the materials used in their construction are suffi-
ciently tissue-equivalent for the intended use.
Some have been designed specifically for radiother-
apy applications and while they scatter and absorb
high-energy photons realistically, this may not be so
for photons of relatively lowenergies used in medical
x-ray imaging (Shrimpton et al., 1981). Require-
ments for physical dosimetry phantoms to be used
in medical x-ray imaging, together with descriptions
of commercially available products, are given in
ICRU Report 48 (1992c).
Reliable estimation of the mean dose to large or
widely distributed organs, like the lungs or the act-
ive BM, requires a large number of measurement
points within the phantom. The absorbed dose is
unlikely to be uniformly distributed throughout the
organ. Physical measurements are best made with
TLDs whose small size allows sufficient numbers
to be suitably distributed. Preparation and read-out
of the TLDs can be very tedious. Realistic hetero-
geneous phantoms simulating bony structure and
soft tissues are very expensive.
International Commission on Radiation Units and Measurements 2005
5.2 MONTE CARLO RADIATION TRANSPORT
CALCULATIONS
If a wide variety of clinical situations or exposure
conditions need to be considered, a mathematical
representation of the patient and theoretical calcu-
lation of D
T
may be more practicable than measure-
ments with a physical anthropomorphic phantom.
Theoretical calculations of D
T
have mostly been per-
formed using the Monte Carlo techniques described
in this section, which simulate the passage of
x-ray photons through the phantom. The paramet-
ers defining the x-ray beam can be easily modified
so that D
T
relative to a normalization quantity, for
example, K
a,i
can be calculated for any desired x-ray
field. However, considerable level of scientific under-
standing is needed to use the codes properly.
5.2.1 Main features of the Monte Carlo
technique
The Monte Carlo technique is a computational
method based on the mathematical simulation of
physical processes. For application of the technique,
the physical processes need to be reformulated in
terms of the probabilities of the variables involved.
Probability functions and scoring functions need to
be defined. Once this is done, the complex problem
reduces to a simpler problem of random selection of
values of each variable from the probability func-
tions. By calculating many cases (histories) reliable
average values can be obtained with an estimate of
their statistical uncertainties.
In patient dosimetry for medical x-ray imaging,
Monte Carlo calculations are used to simulate and
record the energy deposition of x-ray photons in com-
putational models of the human body. The radiation-
interaction histories of a large number of incident
photons are followed using known physical descrip-
tions of the interaction processes and the resulting
energy depositions at the sites of interaction are
recorded. For medical x-ray imaging, the photon
interactions simulated are limited to the photoelec-
tric effect and coherent and incoherent scattering
because the initial photon energies are <150 keV.
The energy given to the secondary electrons is usu-
ally assumed to be absorbed at the point of interac-
tion. Consequently, the kerma is numerically equal
to the absorbed dose. The ranges of the secondary
electrons are small compared with the dimensions of
the relevant organs, and the absorbed dose does not
change abruptly with distance except at a boundary
where composition and density change. In general,
these boundary effects have little impact on the
determination of the average absorbed doses in the
tissues. The single exception is the active BM, where
a small increase in absorbed dose due to the size of
the marrow cavities is expected from increased pho-
toelectron emission by surrounding bone (Spiers,
1963). The bremsstrahlung produced in the tissues
of the human body by secondary electrons is negli-
gible for photons with energies <150 keV.
Organ doses are obtained by summing, in each
organ, all energy depositions from primary and scat-
tered photons, and dividing by the total organ mass.
The result is the mean absorbed dose in the entire
organ regardless of the fraction of the organ irradi-
ated. The standard uncertainty
1
of the mean is
dependent on the number of histories. A history
comprises the path of the primary photon, the sec-
ondary particles it produces in the regions of inter-
est, and the locations and quantities of energy
deposition. According to the central limit theorem
(Feller, 1968), the standard uncertainty of the
mean decreases in inverse proportion to the square
root of the number of primary photons followed, i.e.,
the number of histories. The standard uncertainty of
the mean are computed as a measure of the reliabil-
ity of the results from Monte Carlo calculations.
5.2.2 Main features of the computational
models of the human body
Two distinct types of computational model have
been developed, namely, mathematical and voxel
(volume pixel) phantoms. Both are described in
ICRU Report 48 (1992c).
5.2.2.1 Mathematical phantoms
A three-dimensional mathematical phantom rep-
resenting a reference adult patient was originally
developed at the Oak Ridge National Laboratory,
USA, for use in the field of internal dosimetry (Sny-
der et al., 1969). It was based on the physique and
anatomy of the Reference Man (ICRP, 1975) and
formulae describing its outline and the shape and
disposition of internal structures and organs
appeared in Pamphlet No. 5 of the Medical Internal
Radiation Dose Committee (MIRD).
A revision of the MIRD-5 phantom was made by
Cristy (1980). In addition to including phantoms
representing children of various ages, Cristy uses
an improved model for the distribution of the active
BM, a more sophisticated heart model and adds
regions simulating the female breast to the trunk.
Kramer et al. (1982) developed male and female
adult phantoms, ADAM and EVA, having the appro-
priate gender specific organs and with EVA having
dimensions 0.83 times that of ADAM. A more real-
istic neck region was built into these phantoms to
1
Older texts use standard error.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
56
improve estimates of thyroid doses from anteriorly
incident x rays. Jones and Wall (1985) used a herm-
aphrodite adult phantom based on Cristys adult
phantom with Kramers modified neck region and
breast tissue composed of 50 % fat and 50 % water
rather than the standard soft tissue composition
used in the other phantoms. In the earlier (before
1991) versions of these mathematical models the
oesophagus was missing, because this was only
introduced as a radiosensitive organ in ICRP Pub-
lication 60 (1991b).
All current mathematical phantoms consist of an
elliptical cylinder representing the trunk and arms,
two truncated circular cones representing the legs
and an elliptical cylinder capped by half an ellipsoid
representing the head. The skeleton and organs are
mostly generated from further truncated elliptical
cylinders or ellipsoids specified by quadratic inequal-
ities. Such crude formulations for the organ bound-
aries permit reasonable computational efficiency in
carrying out the Monte Carlo calculations but pro-
duce the limitations in anatomical accuracy that are
evident in Figure 5.1. Exact delineation of the x-ray
beams used in medical imaging becomes more diffi-
cult than with realistic physical phantoms, leading
to greater uncertainty in the doses calculated for
small organs close to beam edges. However, the
mean doses to large organs are probably estimated
more accurately by the Monte Carlo technique where
the total energy deposited in the organ is calculated,
rather than by averaging dose measurements made
at a limited number of points within the organ.
5.2.2.2 Special features of the active BM
In the Monte Carlo procedure, energy deposition is
recorded in each of the sub-regions of the skeleton.
Absorbed dose to the active BM is derived from the
summation of energy depositions in each skeletal
region. The result is the absorbed dose weighted
over the total active BM regardless of the fraction
of the skeleton irradiated. This procedure is consist-
ent with recommendations for partial organ irradi-
ation when absorbed doses are <1 Gy (ICRP, 1969).
Figure 5.1. Various cross sections of the female mathematical phantom EVA (based on data of Kramer et al., 1982). Horizontal
cross sections refer from top to bottom to various levels in the head, the chest region, the abdomen region, and the legs. Reproduced
with permission from Dr. F.W. Schultz.
METHODS FOR DETERMINING ORGAN AND TISSUE DOSES
57
In the phantoms, the bone and marrow are mixed
homogeneously in the skeleton, that is, there is no
geometrical representation of the marrow sites as
separate sites of interaction. The composition and
density of the skeleton is that of a constant admix-
ture of the two components, with each skeletal
region being assigned an appropriate mass of active
BM. The energy deposited in the active BM compon-
ent of this admixture is based on the ratio of the
mass energy absorption coefficient for active marrow
to that of the overall skeletal region, and the fraction
of skeletal mass that is active BM (Rosenstein,
1976).
The increase in absorbed dose to active BM in
small cavities due to the increase in photoelectron
emission by surrounding bone is 220 % for monoen-
ergetic photons in the range 20150 keV. The
increase depends on the photon energy and the size
of the trabecular cavities (Spiers, 1963). The range of
correction factors from one bone to another for a
given energy is less than 4 % of the average value.
Average values were generally used, since x-ray
fields encompass more than one skeletal component.
In the peak energy range found in most x-ray spectra
used in radiology, 3040 keV, the correction is of the
order of 10 % (Rosenstein, 1976).
5.2.2.3 Voxel phantoms
More realistic voxel phantoms are becoming
available based on whole-body CT or magnetic res-
onance imaging scans of real patients. Each organ or
tissue in these phantoms consists of volume ele-
ments derived from the scan data so that tissue
boundaries can be more accurately modelled than
with the simple geometric shapes of the earlier
mathematical phantoms. Spatial resolution is lim-
ited only by the chosen voxel size. So far the results
of Monte Carlo calculations in voxel phantoms for
diagnostic x-ray exposures are available for the
BABY and CHILD phantom (see, Figure 5.2; Zankl
et al., 1988; Veit et al., 1989). Adult voxel phantoms
have been developed (Zankl and Wittmann, 2001;
Petoussi-Henss et al., 2002) and will be used for
patient organ dose calculations.
5.2.3 Uncertainties in Monte Carlo
organdose calculations
Coefficients of variation, for example, relative
standard deviations on the mean, are computed as
a measure of the reliability of the statistical aspects
of the Monte Carlo calculations (Section 5.2.1). Often
more than one million primary photons are followed
to reduce the statistical uncertainties to reasonable
levels. For the calculation of dose-conversion coeffi-
cients, for example, D
T
per unit of K
a,i
, for PA chest
radiography of adults Schultz et al. (1994) followed
in general 10
6
histories. A history comprises the
path of a primary photon, and of the secondary par-
ticles it gives rise to, through the regions of interest,
including the locations and quantities of energy
dissipation. The statistical uncertainty of the dose-
conversion coefficients is largely determined by that
of the computed D
T
, because the relative error in
the K
a,i
is only about 0.1 %. The relative errors in
D
T
vary considerably: (a) for each individual organ,
depending on the exposure parameters (spectrum
and geometry); (b) between organs, depending on
whether an organ is inside or far away from the
Figure 5.2. The exterior surface (top) and selected internal organs
(bottom) of the child phantom (Zankl et al., 1988; Veit et al., 1989,
with permission).
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
58
primary beam, for example, from <0.3 to 0.9 % in
lungs to >19 to 48 % in the lower large intestines, in
the case of chest radiography.
Other sources of uncertainty in Monte Carlo cal-
culations of this type are uncertainties in the basic
attenuation coefficients and inadequacies in the
description of the x-ray source and the patient. The
linear attenuation coefficients are derived from
atomic cross sections and differential scattering
cross sections, the elemental compositions of the tis-
sues and the tissue densities. Hubbell and Seltzer
(1995) quote uncertainties of the order of 5 % for
elemental mass attenuation coefficients <5 keV
and only 2 % up to 10 MeV. Therefore, the elemental
attenuation data should not be a significant source of
error.
The elemental compositions and densities of all
the individual tissue types are not modelled in math-
ematical phantoms. In ADAM and EVA, four tissues
are considered (lung, skeletal, skin, and soft tissue),
and in voxel BABY and CHILD phantoms five and
six tissues, respectively. In more recent voxel phan-
toms, for example, GOLEM (Zankl and Wittmann,
2001), the number of tissues has been extended to
eight. A study comparing organ doses calculated
with phantoms composed of 4 and 18 distinct tissues,
respectively, indicates that the uncertainties made
by considering only a small number of tissues gener-
ally do not exceed 10 % for radiation qualities typical
of diagnostic radiology (Pani et al., 1987).
Of probably greater significance is the impossibil-
ity of describing accurately the juxtaposition of
human anatomy and x-ray beam configurations
with mathematical phantoms. For instance, the
lungs in the mathematical phantoms ADAM and
EVA are not curved at the mediastinum. Con-
sequently, the whole lung can be imaged in a field,
which includes smaller fractions of, for example,
stomach, kidneys, liver, and pancreas than is the
case for real patients. Here the voxel type of human
body phantom offers a clear improvement, because
the organ shapes are replicated from the imaging of
real persons. Relatively small shifts, for example, 5
cm in height for chest radiography, can result in
considerable changes in organ doses, for example, a
factor of 3 for the dose to the spleen (Gosch and
Gursky, 1992; Schultz et al., 1994).
5.2.4 Comparison of conversion
coefficients calculated at different
institutes
For PA chest radiography, a comparison has been
made of the dose-conversion coefficients calculated
by Schultz et al. (1994) with those calculated at GSF
(Drexler et al., 1990), CDRH (Rosenstein, 1988), and
NRPB (Jones and Wall, 1985). For the five organs
receiving the highest doses, the agreement between
the GSF and Schultz et al. (1994) data for the ADAM
phantom is within 10 % when similar vertical beam
position and the same spectra are used. A compar-
ison of conversion coefficients at 100 kV for the male
phantoms for lung and active BM between CDRH
and Schultz et al. shows excellent agreement when
the same vertical field positions are used (Schultz
et al., 1994). However, for the thyroid the differences
are larger (up to 50 %) and strongly dependent on
the vertical beam position. The comparison of the
Schultz et al. and NRPB data is complicated by the
fact that NRPB results refer to a hermaphrodite
phantom. The differences between the conversion
coefficients for ADAM and the hermaphrodite phan-
tom (NRPB) remain generally within 20 % for the
five most heavily irradiated tissues.
For CT, the data published by the NRPB (Jones
and Shrimpton, 1993) for the Philips Tomoscan 350
scanner showed on average agreement within 10 %
with the data calculated for a mimicked hermaph-
rodite phantom by Jansen et al. (1995).
The data calculated by PCXMC (Tapiovaara et al.,
1997; Appendix F) are most directly comparable
with the organ dose-conversion coefficients calcu-
lated at the NRPB by Jones and Wall (1985) and
Hart et al. (1994b, 1996b), because they are based
on the same phantom models. In general, the agree-
ment between the results of PCXMC and the NRPB
data is good. The differences between them typically
fall within the statistical uncertainty and are
generally <10 %.
5.2.5 Comparison of measured and
calculated organ doses
A number of experimental comparisons were
made during the original work with the Monte
Carlo code (see Appendix B.2.1) at the US-Bureau
of Radiological Health (now the Center for Devices
and Radiological Health) by Rosenstein (1976), Beck
and Rosenstein (1979), and Chen et al. (1978). The
results of these comparisons are summarized below.
In addition, a comparison of measured organ doses
with values derived using calculated organ dose-
conversion coefficients is presented for CT (Geleijns
et al., 1994).
5.2.5.1 Adult phantoms: organs in the x-ray field
The central axis depth doses available at the time
for use in radiotherapy (Cohen et al., 1972) could be
used for dose assessment when small organs such
as the ovaries or testes are located centrally in a
radiation field (Rosenstein, 1976). The irradiation
conditions are comparable with those for the ovaries
METHODS FOR DETERMINING ORGAN AND TISSUE DOSES
59
and testes in the Monte Carlo phantom for the AP
view, where the organs are surrounded by tissue and
not in direct contact with nor shielded by portions of
the skeleton. Additional comparisons can be made
for the thyroid in the AP view and the ovaries in
the PA view, but in these cases the inhomogeneities
in the Monte Carlo phantom affect the radiation
transport.
An additional experiment for absorbed dose to the
testes used an adult Alderson phantom (McDowell
et al., 1976). The Alderson phantom was constructed
of tissue-equivalent material, contained a full skel-
eton and simulated lungs, and had anthropometric
characteristics that closely matched the reference
man dimensions of the Monte Carlo phantom.
The agreement among the three sets of data for
the testes (AP view) was within 15 % in all cases and
generally within 3 %. Agreement between two sets of
data for the ovaries (AP view) was within 20 % in all
cases and generally within 11 % (Rosenstein, 1976).
5.2.5.2 Adult phantoms: organs outside
the x-ray field
The experiment for absorbed dose to the testes
noted above included several irradiations in which
the testes were in close proximity (i.e., testes center
just within the field edge) or outside the x-ray field
(i.e., testes center 3, 5, or 9 cmoutside the field edge).
This allowed a check of the radiation transport
calculation when scattered radiation conditions
prevailed. Agreement of the two sets of data was
generally within 30 % (Rosenstein, 1976).
In the same experiment (Rosenstein, 1976), irra-
diations were made with and without a lead shield
placed over the simulated testes to prevent external
photon scatter from the x-ray filter and intervening
air column to the testes. Such scatter is ignored in
the Monte Carlo calculation. Without the shield, the
absorbed doses to the testes were 2550 % higher
than with the shield. This external scatter contribu-
tion is <1 % of the total dose when the testes are in
the x-ray field.
5.2.5.3 Adult phantoms: active BM
Data from the two BM dosimetry approaches
available at the time (Ellis et al., 1975; Rosenstein,
1976) were not directly comparable because of differ-
ent anatomical and technical assumptions, but a
comparison was instructive (Kereiakes and Rosen-
stein, 1980). Even the same HVL
1
does not yield the
same results. When the same spectra are used the
differences are smaller.
The more significant differences in the two
approaches are described in Table 5.1.
For identical x-ray fields and HVL
1
, the Ellis et al.
(1975) absorbed dose values were up to a factor of
23 higher. When the same beam quality from the
experimental approach was used in the Monte Carlo
code, the absorbed doses were a factor of 1.21.8
times higher than those obtained with the corres-
ponding softer beam qualities. The effect of the dif-
ferences in lung models was to increase the active
BM dose at those sites directly behind the lungs by
up to 25 %. These results accounted for much of the
difference obtained between the two approaches.
5.2.5.4 Paediatric phantoms: head and neck
A tissue-equivalent head and neck phantom
incorporating a skull of a child 6 years old was
used to obtain experimental absorbed doses to the
thyroid and brain for comparison to the results for
the Monte Carlo phantom for the 5-year-old child
(Beck and Rosenstein, 1979). The statistical uncer-
tainty (at 2 SD) associated with the experimental
data was <20 %; the statistical uncertainty of the
computed data was <3 % for brain doses and <10 %
for thyroid doses. For the brain doses, the ratio of the
experimental to computed doses ranged from 0.66 to
0.94; for the thyroid doses, the ratios ranged from
1.13 to 1.21.
5.2.5.5 Paediatric phantoms: whole body
Tissue-equivalent phantoms that were nearly
identical to the mathematical child phantoms in
the Monte Carlo code were constructed for 1 and 5
year olds. Measurements of absorbed doses were
made at sites for the brain, each lobe of the thyroid,
each ovary, the testes, and several sections of the
skeleton, i.e., the spine, each upper arm, the pelvis,
and each upper leg (Chen et al., 1978). Ratios of the
experimental to computed results varied from 0.61
to 1.39. Most were between 0.75 and 1.25 and no
Table 5.1. The more significant differences in bone dosimetry approaches of Ellis et al. (1975) and of Rosenstein (1976).
Experimental approach (Ellis et a1., 1975) Monte Carlo approach (Rosenstein, 1976)
Inflated lung model: Relaxed lung model:
Volume and density 5000 cm
3
; 0.2 g/cm
3
3378 cm
3
; 0.3 g/cm
3
Beam quality: Beam quality:
e.g., HVL
1
2.3 mm Al e.g., HVL
1
2.3 mm Al
(Peak tube voltage: 105 kV, 1 mm Al added filtration) (Peak tube voltage: 75 kV, 2.7 mm Al total filtration)
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
60
trend could be discerned for those comparisons lying
outside these limits.
5.2.5.6 Adult phantoms: CT
Anthropomorphic body phantoms have been used
for dosimetry in diagnostic radiology (see, e.g.,
Maccia et al., 1988; Geleijns et al., 1992, 1994).
Figure 5.3 shows the results of an experimental
method and two computational methods for the
determination of organ doses for CT abdomen invest-
igations (Geleijns et al., 1994). Differences in organ
dose are observed for the different methods. It is
expected that these variations are mainly due to
differences between the phantoms used. When com-
paring the mathematical phantoms of the NRPB
(Jones and Wall, 1985) and GSF (Kramer et al.,
1982), it is noticed that they have very similar dis-
tribution of the radiosensitive organs. In the study of
Geleijns et al. (1994) an anthropomorphic RANDO
phantom was used to estimate organ and tissue
doses. The RANDO phantom is different in size com-
pared with the mathematical phantoms employed by
NRPB and GSF and the estimated location of the
organs in the RANDO phantom differs significantly
in comparison with the distribution of organs in
the mathematical phantom. This explains the large
difference between the measured and calculated
values for the dose to the bladder.
5.2.5.7 Sources of data on dose-conversion
coefficients
Important features of published tabulations of
coefficients for relating measurable normalization
quantities to organ doses are shown in Table 5.2.
In addition, Table 5.2 provides information on the
application of the Monte Carlo code PCXMC for cal-
culating dose-conversion coefficients. All conversion
coefficients rely on Monte Carlo simulation of the
radiation transport through mathematical or voxel
phantoms.
More nearly complete references to these sources
of dose-conversion coefficients with details of their
important features are given in Appendices BD,
grouped according to the organization where they
were developed, and in Appendices E and F
Appendices BD include sample pages from the
tabulations and example calculations of organ
doses for given x-ray examination conditions.
The information in Appendices BD is extensive
but not comprehensive. The variation in exposure
conditions for a particular examination or procedure
is large (radiation quality, field size, and location on
the body, view) and the number of types of examina-
tions and procedures is large and varying. Often
practice in one country is different from that in
another country. Appendices B, C, and D reflect
practice in the US, Germany, and the UK, respect-
ively. Differences in examinations or procedures can
even exist between radiologists in the same hospital.
It is, therefore, impossible to make comprehensive
calculations.
Acomparison of organ dose-conversion coefficients
calculated at different institutes reveals that for
similar exposure conditions the values for the
internal organs located in the primary beam show
agreement within 10 % (Section 5.2.4). For similar
exposure conditions, comparisons of calculated and
measured D
T
also show acceptable agreement
(Section 5.2.5).
Differences in exposure conditions result in larger
differences in D
T
for the same type of medical
Figure 5.3. Organ doses from CT abdomen examinations
determined using TLD dosimetry and a RANDO phantom
(black) or using C
K
values and organ dose-conversion coefficients
of the NRPB (gray) and GSF (white) (Reproduced from Geleijns
et al., 1994, with permission of British Institute of Radiology).
RBM is red BM.
Appendix B Handbooks produced by the Center
for Devices and Radiological Health
(CDRH)
Appendix C Reports produced by the German
Center for Health and the
Environment (GSF)
Appendix D Reports produced by the Health
Protection Agency (HPA) (Formerly
National Radiological Protection
Board NRPB)
Appendix E Review of Monte Carlo calculations
for assessment of mean glandular
dose in mammography
Appendix F PCXMC A PC-based Monte Carlo
program for calculating patient dose
due to radiology
METHODS FOR DETERMINING ORGAN AND TISSUE DOSES
61
imaging task (Wise et al., 1999). When for the PA
chest examination the field is shifted vertically by
5 cm, differences of up to 60 %were found in organs
partly irradiated by the primary beam (Gosch and
Gursky, 1992; Schultz et al., 1994). For this exam-
ination differences in x-ray spectra, i.e., 130 kV
x rays versus 80 kV x rays, result in conversion
coefficients for the most heavily irradiated organ
(lungs) differing by a factor of 2 (Schultz et al.,
1994). In the publication by Drexler et al. (1990)
organ dose coefficients are given for various x-ray
qualities (three values of tube voltage per type of
examination) for a series of common x-ray projec-
tions. For organs in the centre of the field, differences
of up to a factor of 2 were found; for organs at
the edge or outside the field the variations can be
larger.
When a dose-conversion coefficient is needed for a
specific situation, it is recommended to select a value
from the available data based on similarities in
exposure conditions (projection, view, field size, and
radiation quality) and patient model.
Table 5.2. Important features of sources of dose-conversion coefficients for medical x-ray imaging.
Type of examination No. of
views
No. of
organs
No. of
spectra
Normalization
quantity
Phantom Reference
a
General radiology 54 8 17 K
a,i
MIRD hermaphrodite CDRH 89-8031 (Rosenstein, 1988)
40 16 3
b
K
a,i
ADAM EVA GSF 11/90 (Drexler et al., 1990)
68 26 40 K
a,e
, P
KA
Cristy hermaphrodite NRPB-SR262 (Hart et al., 1994b)
c
24
c
K
a,i
, P
KA
Cristy hermaphrodite
c
Tapiovaara et al. (1997)
Mammography 6 1 14 K
a,i
Reference CDRH 85-8239 (Rosenstein et al., 1985)
1 1 90 K
a,i
Breasts Dance (1990, 2000)
1 1 10 K
a,i
Wu et al. (1991a, 1994)
1 1 17 K
a,i
Jansen et al. (1994)
Fluoroscopy: 12 12 3 K
a,i
ADAM CDRH 92-8282 (Rosenstein et al., 1992)
Upper GI tract EVA
Coronary 20
d
6 K
a,i
ADAM CDRH 95-8289 (Stern et al., 1995b)
Arteries EVA
Fluoroscopy
c
24
c
K
a,i
, P
KA
Cristy hermaphrodite
c
Tapiovaara et al. (1997)
Pediatric
radiography
20 6 3 K
a,I
0, 1, 5 years
hermaphrodite
BRH 79-8079 (Rosenstein et al., 1979)
20 26 72 K
a,i
, P
KA
0,1,5,10,15 years NRPB-R279 (Hart et al., 1996a)
hermaphrodite NRPB-SR279 (Hart et al., 1996b)
5/6 16/11 1
b
K
a,i
Voxel baby Zankl et al. (1988)
Voxel child Zankl et al. (1989)
Radiography/
Fluoroscopy
c
24
c
K
a,i
, P
KA
Christy
hermaphrodite
c
Tapiovaara et al. (1997)
CT adult 208 23 23 C
K
e
Cristy NRPB-R250 (Jones and Shrimpton, 1991)
slices hermaphrodite NRPB-SR250 (Jones and Shrimpton, 1993)
104 slices 22 3 C
K
ADAM EVA GSF 30/91 (Zankl et al., 1991)
CT Pediatric 45 slices 35 2 C
K
Voxel baby GSF 30/93 (Zankl et al., 1993)
66 slices 37 2 C
K
Voxel child
a
For more complete information see Appendices BF.
b
Per view.
c
Views, spectra and phantom sizes can be freely selected.
d
Data are also given for the skin in the primary field (maximum dose).
e
Expressed in muscle tissue instead of air.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
62
Journal of the ICRU Vol 5 No 2 (2005) Report 74 doi:10.1093/jicru/ndi028
Oxford University Press
6 CONCLUSIONS
The two purposes of patient dosimetry for medical
x-ray imaging are to set and check standards of good
practice and to assist in assessing detriment or
harm. The present situation in patient dosimetry
for medical x-ray imaging clearly indicates the
needs for dose quantities recommended for different
applications and for using the same, self-consistent
system for names, symbols, and units.
The radiation quality of an x-ray beam can be
characterized by the x-ray spectrum. Although x-
ray spectra can be measured, it is recommended
that the radiation quality of x-ray beams used for
medical imaging be characterized by a combination
of various parameters. These include the first HVL,
HVL
1
, the second, HVL, HVL
2
, the ratio of HVL
1
to
HVL
2
, referred to as the homogeneity coefficient, the
tube voltage and the total filtration. In most cases,
the quality of an x-ray beam can be adequately spe-
cified by means of the combined information on tube
voltage, HVL
1
, and HVL
2
, or the tube voltage, HVL
1
and total filtration. The x-ray tube output is also an
important characteristic of an x-ray tube (including
filtration).
Several application-specific dosimetric quantities
have been found useful for measurements in medical
x-ray imaging. However, ambiguity exists in names
of quantities and their use. Because of the equival-
ence of numerical values of absorbed dose and kerma
in the same material for the x-ray energies used in
medical imaging, quantities have often been altern-
atively referred to in terms of absorbed dose or in
terms of kerma. Because almost all of these quantit-
ies will be measured with instruments calibrated
free-in-air in terms of air kerma, it is considered to
be appropriate and more consistent with practice
elsewhere to name them in terms of air kerma.
Recommended application-specific quantities for
dosimetry in medical x-ray imaging are summarized
in Table 3.1 and followed by more detailed
definitions.
At low dose levels, the mean absorbed doses in
organs or tissues in the human body are taken to
be an indication of the probability of subsequent
stochastic effects. At high dose levels, the absorbed
dose to the more heavily irradiated sites within the
body is taken to be an indicator of the severity of
deterministic effects. For an assessment of the risk
due to induction of stochastic and deterministic
effects by medical x-ray imaging detailed knowledge
is required of organ doses, absorbed dose distribu-
tion, and the age and gender of the group of patients
concerned, rather than effective dose.
A dose-conversion coefficient relates a specified
dosimetric quantity to some other quantity, i.e., the
normalization quantity, which can be readily
measured or calculated in the clinical situation.
For stochastic effects, the specified dosimetric
quantities are the organ doses, orin the case of
mammographythe mean glandular dose. Normal-
ization quantities are K
a,i
, K
a,e
, P
KA
, or C
K
. In viewof
the technical developments in CT, particularly large
volume scanning, basic measurements using a 10 cm
long detector (ionization chamber) will be most prob-
ably inadequate in the future. For interventional
radiology, dose-conversion coefficients with the aim
to avoid the occurrence of deterministic effects are
lacking until now except for some coronary proced-
ures (Appendix B).
Diagnostic reference levels, DRLs, are introduced
as a form of investigation level in diagnostic radi-
ology. DRLs are aimed at the management of patient
doses consistent with the clinical information that is
required. The dosimetric quantities recommended
for the establishment of DRLs are K
a,i
, K
a,e
, D
G
,
and P
KA
.
Methods are required for measurement of normal-
ization quantities used in the dose-conversion coef-
ficients (Section 5 and Appendices BF) and of
quantities recommended for establishment of
DRLs. Measurement of these quantities mainly
relies on the use of ionization chambers and solid-
state dosimeters, including TLDs. Calibration pro-
cedures including requirement for stating uncer-
tainties are considered essential. The air kerma
free-in-air is the desired dosimetric quantity for the
x rays used in medical imaging. The uncertainty of
dose measurements for comparative risk assessment
as well as for quality assurance should not exceed
about 7 % using a coverage factor of 2. This implies
that well-defined conditions for the calibration at
International Commission on Radiation Units and Measurements 2005
standards laboratories must be established. In a few
cases calibration of dosimeters at a standards labor-
atory is neither feasible nor possible. This includes
air kermaarea product meters that should be calib-
rated in situ and TLDs that should be calibrated
based on the actual field conditions. For these types
of dosimeters specific information is provided.
Examples are given of radiation qualities used for
calibration of dosimeters in conventional radiology.
An international code of practice for patient dosi-
metry in diagnostic radiology is presently being
developed by the IAEA. It will include practical
details of calibrations at the standards laboratory.
Information is provided for measurement of
application-specific dosimetric quantities including
incident air kerma K
a.i
, entrance surface air kerma
K
a.e
, air kermaarea product, P
KA
, especially during
complex examinations involving both fluoroscopy
and radiography, and dosimetric quantities used in
CT. In this respect it should be borne in mind that
the quantities recommended for dosimetry in CT are
of a provisional nature.
Methods of determining organ and tissue doses
are summarized in Section 5. Dose measurements
in physical anthropomorphic phantoms have been
applied for this purpose but can be very tedious,
and realistic physical phantoms are very expensive.
Therefore, Monte Carlo methods are commonly used
to derive dose-conversion coefficients relating nor-
malization quantities to mean organ and tissue
doses, as well as doses to localized areas of, for
example, skin. In addition, general information is
provided on Monte Carlo methods used for computa-
tion of radiation transport in anthropomorphic
mathematical phantoms or anthropomorphic voxel
phantoms. A comparison is made of conversion coef-
ficients calculated at various institutes and between
measured and calculated organ doses.
Appendix A includes specific information on
backscatter factors commonly required for dosi-
metry of low, for example, mammography, and
medium energy, and all other energies of x rays
used in medical imaging. Backscatter factors are
dependent on radiation quality, for example, HVL,
field size and to a lesser extent on the x-ray spec-
trum, target angle, focal spot-to-surface distance,
and phantom thickness and material.
Dose-conversion coefficients for specific proced-
ures applied for x rays used in medical imaging are
presented in Appendices BE. A comparison of con-
version coefficients calculated at different institutes
shows that differences are small, i.e., within the
statistical errors and generally less than 10 %
when the exposure conditions and mathematical
models of patients are similar. Larger differences
occur due to differences in mathematical models
simulating the patient. Similar conclusions can be
drawn fromcomparisons of measured and calculated
organ dose-conversion coefficients. Larger errors,
i.e., well in excess of 10 % occur for conversion coef-
ficients concerning organs far away from the prim-
ary beam. This is not a practical problem since their
contribution to the radiation risk is much less than
that of organs receiving the highest doses.
When a dose-conversion coefficient is needed for
a specific situation, the best approach is to select
a value from the available data based on similari-
ties in exposure conditions (projection, view, field
size, and radiation quality) and patient model.
Appendices B, C, and D reflect practice in the
USA, Germany, and the UK, respectively, whereas
Appendix E deals with mammography. Appendix F
provides information on a PC-based Monte Carlo
program for calculating patient doses in radio-
graphy, excluding mammography, and fluoroscopy.
It also allows variations in patient size.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
64
Journal of the ICRU Vol 5 No 2 (2005) Report 74 doi:10.1093/jicru/ndi029
Oxford University Press
APPENDIX A: BACKSCATTER FACTORS
Various sources of backscatter factors B for low-
energy x rays incident on homogeneous phantoms
are available in the literature. Some have been
derived for exposure conditions typical of radio-
therapy practice (Chan and Doi, 1981; BJR, 1983;
Harrison et al., 1990; Klevenhagen, 1982, 1989;
Grosswendt, 1984, 1990, 1993; IPSM, 1991; Ma
et al., 2001) and others for exposure conditions typi-
cal of diagnostic radiology (Stanton et al., 1982;
Harrison, 1982; Petoussi-Henss et al., 1998). In addi-
tion, backscatter factors have been calculated using
Monte Carlo techniques for diagnostic x-ray fields
incident on anthropomorphic mathematical phan-
toms (computational models) of the whole body, con-
taining simulated bone and lung structures within a
soft tissue-equivalent material (Hart et al., 1994a,b).
It should be noted that different definitions of the
backscatter factor B are used. For radiotherapy
applications, B is commonly defined as the ratio of
the water kerma at a point on the beam axis at the
surface of a sufficiently large (water) phantom to the
water kerma at the same point in the primary (i.e.,
incident) beam, with no phantom present. In alter-
native definitions the ratios of air kermas or of
absorbed doses, for example, IAEA (1987) are used.
For the x-ray energy range typically used in radiol-
ogy, the quantities water kerma and absorbed dose
to water have effectively the same values.
Values of B are primarily a function of the x-ray
spectrum (HVL
1
) and field size but are also depen-
dent to a lesser extent on the anode angle, focus-to-
surface distance, d
FSD
, and phantom thickness and
material. Measurements at diagnostic radiology con-
ditions by Harrison (1982) showed that for a given
HVL
1
, the value of B is dependent upon the peak
tube voltage. Qualitatively, for a given HVL
1
a
higher peak tube voltage implies a broader spectrum
leading to a B modified by lower and higher energy
scatter. The effect reported by Harrison (1982)
amounts to 2% at maximum at a HVL
1
of 2 mm
Al and tube voltages ranging from 60 to 100 kV.
The influence of the anode angle on B has been
qualitatively investigated by Carlsson (1993). The
smaller the anode angle the larger is the heel effect,
i.e., the lower is the fluence rate at the anode side of
the field. Furthermore, the smaller the anode angle
the more the anode material acts as a filter. This
latter effect, however, also influences the HVL
1
.
For large beam angles and a small anode angle, the
heel effect reduces the value of B. For radiology
(except mammography), it might be expected that,
because of the relatively large d
FSD
the influence of
anode angle on the value of B is most likely a few
percentage at maximum.
Calculations of Grosswendt (1990) demonstrate
an increase in B with increasing d
FSD
. For values of
d
FSD
in the range from 30 to 100 cm the variation in
B at a field size of 20 cm diameter is 3% at max-
imum. For values of d
FSD
>100 cm, the change in
B with d
FSD
will be even smaller. It is concluded
that the influence of changes in d
FSD
on the value
of B will generally be negligible for radiology
applications.
The thickness of the underlying phantom material
influences the magnitude of the backscatter factor.
Klevenhagen (1982) presents a set of equations
allowing the calculation of the backscatter factor
for given field size, HVL
1
,
,
d
FSD
, and thickness of
underlying material. For practical purposes in radi-
ology, it is concluded that a phantom thickness of 8
cm is sufficient to obtain a backscatter factor >95%
of the value with full backscatter for all field sizes
and radiation qualities. This is in agreement with
the conclusions of Stanton et al. (1982).
The elemental composition of a phantom also
influences B. Compared with a water phantom,
the use of tissue substitute materials generally
result in small, i.e., >5% corrections, dependent
on the radiation quality (Grosswendt, 1984). For
polymethylmethacrylate (PMMA), however, devia-
tions >5% occur at a field size of 100 cm
2
and a
radiation quality characterized by a HVL
1
of 2 mm
Al. This is in agreement with findings of Stanton
et al. (1982) for measurements with PMMA, BR12
(average breast tissue substitute), and MS11 (water
substitute).
Backscatter factors for diagnostic radiology were
calculated by Petoussi-Henss et al. (1998) using
Monte Carlo methods. Values of B have been
obtained for cuboid (30 cm 30 cm surface and
International Commission on Radiation Units and Measurements 2005
15 cm depth) phantoms of water, PMMA, and ICRU
tissue (elemental composition by mass 10.1% H,
11.1% C, 2.6% N, and 76.2 % O) at different field
sizes and d
FSD
for monoenergetic photons (range
0.011 MeV) and various x-ray spectra
(Table A.1). The dependence of B on photon energy
and field size is significant. The results show that
PMMA is not tissue-equivalent in this energy range.
Backscatter factors have also been derived by
Hart et al. (1994a,b) by calculating the energy
absorbed in a small volume of material on the
entrance surface of an anthropomorphic phantom
that contains bone and lung structures, to allow for
the influence of underlying bone or soft tissue in the
x-ray field. Bone can lie close to the surface where
dosimeters may be attached for measuring entrance
surface dose, for example, for skull radiographs or
PA projections through the chest. Table A.2 shows
backscatter factors for several x-ray projections,
field sizes, and beam qualities. Whereas there is
close agreement with corresponding values in
Table A.1 when the surface dosimeter overlies soft
tissue, there is a reduction in the backscatter factor
by 57% when the point of measurement overlies
lungs (PA chest).
For mammography, the situation is less complex
than for general radiology because the field size, x-
ray spectra, anode angle, focus-to-surface distance,
and phantom thickness and material are less vari-
able. Common conditions in mammography are that
the field size is larger than the phantom (or the
female breast), that a limited number of spectra
Table A.1. Backscatter factors for water, ICRU tissue, and PMMA for some x-ray beams typical of diagnostic radiology at
three field sizes and a focus-to-surface distance d
FSD
of 100 cm (Reproduced from Petoussi-Henss et al., 1998 with
permission from the Institute of Physics).
Tube
voltage/kV
Filter HVL/
mm Al
Field 10 cm 10 cm Field 20 cm 20 cm Field 25 cm 25 cm
Water ICRU tissue PMMA Water ICRU tissue PMMA Water ICRU tissue PMMA
50 2.5 mm Al 1.74 1.24 1.25 1.33 1.26 1.27 1.36 1.26 1.28 1.36
60 2.5 mm Al 2.08 1.28 1.28 1.36 1.31 1.32 1.41 1.31 1.32 1.42
70 2.5 mm Al 2.41 1.30 1.31 1.39 1.34 1.36 1.45 1.35 1.36 1.46
70 3.0 mm Al 2.64 1.32 1.32 1.40 1.36 1.37 1.47 1.36 1.38 1.48
70 3.0 mm Al 0.1 mm Cu 3.96 1.38 1.39 1.48 1.45 1.47 1.58 1.46 1.47 1.59
80 2.5 mm Al 2.78 1.32 1.33 1.41 1.37 1.39 1.48 1.38 1.39 1.50
80 3.0 mm A1 3.04 1.34 1.34 1.42 1.39 1.40 1.51 1.40 1.41 1.52
80 3.0 mm A1 0.1 mm Cu 4.55 1.40 1.40 1.49 1.48 1.50 1.61 1.49 1.51 1.63
90 2.5 mm Al 3.17 1.34 1.34 1.43 1.40 1.41 1.51 1.41 1.42 1.53
90 3.0 mm Al 3.45 1.35 1.36 1.44 1.42 1.43 1.53 1.42 1.44 1.55
90 3.0 mm Al 0.1 mm Cu 5.12 1.41 1.41 1.50 1.50 1.51 1.62 1.51 1.53 1.65
100 2.5 mm Al 3.24 1.34 1.34 1.42 1.40 1.41 1.51 1.41 1.42 1.53
100 3.0 mm Al 3.88 1.36 1.37 1.45 1.44 1.45 1.55 1.45 1.46 1.57
100 3.0 mm Al 0.1 mm Cu 5.65 1.41 1.42 1.50 1.51 1.53 1.64 1.53 1.55 1.66
110 2.5 mm A1 3.59 1.35 1.35 1.43 1.42 1.43 1.53 1.43 1.44 1.55
120 3.0 mm Al 4.73 1.37 1.38 1.46 1.46 1.48 1.58 1.48 1.49 1.60
120 3.0 mm Al 0.1 mm Cu 6.62 1.41 1.42 1.50 1.53 1.54 1.64 1.54 1.56 1.67
130 2.5 mm Al 4.32 1.36 1.36 1.44 1.44 1.45 1.55 1.45 1.47 1.57
150 2.5 mm Al 4.79 1.36 1.36 1.44 1.45 1.46 1.55 1.46 1.48 1.58
150 3.0 mm Al 6.80 1.39 1.39 1.47 1.50 1.51 1.61 1.52 1.53 1.63
150 3.0 mm Al 0.1 mm Cu 8.50 1.40 1.41 1.48 1.53 1.54 1.64 1.55 1.57 1.67
Table A.2. Backscatter factors calculated by Monte Carlo techniques in an anthropomorphic phantom (Hart et al., 1994a;
1994b).
HVL
1
/mm Al Peak tube voltage/kV Total filtration/mm Al Projection
Lateral LSJ
a
AP Abdomen PA Chest
11 cm 14 cm 26 cm 35cm 30 cm 38cm
2.0 60 2.5 1.23 1.31 1.23
2.5 80 2.0 1.25 1.37 1.27
3.0 80 3.0 1.27 1.41 1.30
4.0 110 2.5 1.29 1.45 1.34
a
LSJ is lumbosacral joint.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
66
and anode angles are applied, the focus-to-film
distances are generally 60 cm, that the phantoms
are constructed fromPMMA or BR12. The phantoms
have generally sufficient thickness (i.e., in excess of
3 cm) to approximate full backscatter (Stanton et al.,
1982). Backscatter factors calculated for mammo-
graphy are given in EC (1996b) and in Kramer et al.
(2001).
Table A.3 gives B for the most common spectra in
mammography. These values of B were calculated
(Kramer et al., 2001) for a semi-circular PMMA
phantom of 10 cm radius and 5 cm thickness; focus-
to-skin distance was taken to be 65 cm and the field
on the phantom surface was semi-circular with a
10 cm radius.
The impact of the variation of the exposure
parameters on B was also investigated by Kramer
et al. (2001). It was found that for field radii from 8
to 10 cm, d
FSD
from 60 to 70 cm, phantom thick-
nesses from 3 to 8 cm, and for a rectangular instead
of a semi-circular shape of phantom, this was
negligible. A PMMA phantom, a breast phantom
made of an homogeneous mixture of adipose and
glandular tissue and a phantom consisting of a 3
cm thick homogeneous mixture of adipose and
glandular tissue sandwiched by two 1 cm layers of
adipose tissue, all resulted in practically the same
value of B.
It can be concluded from the table that the varia-
tion in B for mammography is relatively small, that
is, 0.03 from the average value of 1.11. The depen-
dence on phantom material is <5% between PMMA,
BR12, fat, and water (Doi and Chan, 1980; Stanton
et al., 1982).
Table A.3. Backscatter factors for a semi-circular PMMA phantom of 10 cm radius and 5 cm thickness, for three
mammographic spectra; d
FSD
: 65 cm; semi-circular field with 10 cmradius at the phantomsurface (based on data of Kramer
et al., 2001).
Anode: Molybdenum Anode: Molybdenum Anode: Rhodium
Filtration: 30 mm Molybdenum Filtration: 25 mm Rhodium Filtration: 25 mm Rhodium
Peak tube
voltage/kV
HVL
1
/mm Al B Peak tube
voltage/kV
HVL
1
/mm Al B Peak tube
voltage/kV
HVL
1
/mm Al B
24 0.282 1.08 24 0.328 1.09 24 0.297 1.09
25 0.295 1.09 25 0.343 1.09 25 0.316 1.09
26 0.306 1.09 26 0.356 1.10 26 0.330 1.10
28 0.325 1.09 28 0.375 1.10 28 0.362 1.11
30 0.340 1.10 30 0.391 1.11 30 0.391 1.11
32 0.353 1.10 32 0.403 1.11 32 0.418 1.12
34 0.364 1.10 34 0.412 1.11 34 0.441 1.12
35 0.369 1.11 35 0.416 1.12 35 0.452 1.13
36 0.373 1.10 36 0.420 1.11 36 0.462 1.13
38 0.381 1.11 38 0.427 1.12 38 0.481 1.13
40 0.388 1.10 40 0.433 1.12 40 0.497 1.14
APPENDIX A
67
Journal of the ICRU Vol 5 No 2 (2005) Report 74 doi:10.1093/jicru/ndi030
Oxford University Press
APPENDIX B: HANDBOOKS PRODUCED BY THE
CENTER FOR DEVICES AND RADIOLOGICAL
HEALTH (CDRH)
B.1 GENERAL DESCRIPTION
The US Center for Devices and Radiological
Health, Rockville, Maryland (www.fda.gov/cdrh)
has produced five Handbooks that permit the
estimation of typical absorbed doses to tissues of
reference adult and paediatric patients from radio-
graphic and fluoroscopic procedures. The Handbooks
also permit evaluation of the effect on tissue doses of
changes in technical parameters used in or among
facilities. Variations from a particular reference
patient due to anthropometric characteristics of
the human body and its internal organs are not
addressed. Therefore, assignment of organ doses to
individual patients is not recommended.
The specific Monte Carlo codes and anthropo-
morphic phantoms varied with the objectives of
each Handbook and the state-of-the art at the time
each Handbook was prepared. The following five
Handbooks are available.
1. Projections Common in Diagnostic Radiology
(Adults) CDRH 89-8031 (Rosenstein, 1988).
2. Projections Common in Pediatric Radiology
BRH 79-8079 (Rosenstein et al., 1979).
3. Mammography CDRH 85-8239 (Rosenstein
et al., 1985).
4. Upper Gastrointestinal Fluoroscopic Examina-
tions CDRH 92-8282 (Rosenstein et al., 1992).
5. Fluoroscopic and Cineangiographic Examinations
of the Coronary Arteries CDRH 95-8289 (Stern
et al., 1995b).
The earlier Handbooks present tissue doses per
entrance skin exposure (free-in-air)
(D
T
/mrad)/(X
i
/R). Any entry can be converted to
SI units as follows:
To obtain tissue doses per incident air kerma
(D
T
/mGy)/(K
a,i
/mGy) divide the entry by 876.
To obtain tissue doses per incident air kerma
(D
T
/mGy)/(K
a,i
/Gy) divide the entry by 0.876.
The choice of the conversion coefficient depends
on the relative magnitude of the K
a,i
value one would
encounter in a particular radiographic or fluoro-
scopic procedure. The sample tabulations presented
in the present ICRU Report have been converted to
SI units as appropriate for the specific tabulation.
B.2 HANDBOOK CDRH 89-8031
(ROSENSTEIN, 1988): PROJECTIONS
COMMON IN DIAGNOSTIC RADIOLOGY
(ADULTS)
B.2.1 Main features of calculations
The details of the Monte Carlo procedure and
the anthropomorphic phantom used to produce the
Handbook are described by Rosenstein (1976). Sig-
nificant features are the BRHGAM Monte Carlo
code (Warner, 1973), the hermaphrodite MIRD-5
phantom of the Medical Internal Radiation Dose
Committee (Snyder et al., 1969) with supplemental
calculations performed for an added simulation of
the female breasts (Peterson and Rosenstein, 1989),
the elemental compositions for skeletal, lung, and
soft tissue of Tipton et al. (1966), the elemental com-
position of active BM of Poll et al. (1972), the distri-
bution of active BM in the skeletal components of
Ellis (1961), and the photoelectron enhancement
factors for absorbed dose in trabecular spaces of
Spiers (1963).
Except for the female breasts, tissue-air-ratios
were computed for the tissues of interest for monoen-
ergetic photons in the range 20100 keV for 4 cm
4 cm plane-parallel beams normally incident on the
respective mid-plane of the phantom for AP, PA, and
right lateral incidence. The tissue-air-ratios are for
the particular reference phantom. The coefficients
of variation associated with the tissue-air-ratios are
provided in Rosenstein (1976). This approach was
chosen to accommodate the very large number of
International Commission on Radiation Units and Measurements 2005
possibilities for adult projections. Supplemental
tissue-air-ratios were obtained at a later date for
the addition of female breasts for single fields that
encompassed the breasts.
The tissue-air-ratios were then converted to D
T
per
unit K
a,i
coefficients for a specific x-ray projection
using the pertinent information on x-ray spectra,
projection geometries, and field sizes and locations.
The tissue-air-ratios for 4 cm 4 cm grid elements
included in the x-ray field and for photon energies
included in the x-ray spectrum are combined to
simulate the desired conditions. A computer pro-
gram, which performs the conversions of tissue-air-
ratios to tissue doses for common radiographic
projections and simple fluoroscopic projections, and
its documentation are available (Peterson and
Rosenstein, 1989).
B.2.2 Scope of the handbook
The Handbook presents a series of 54 tabulations
of tissue dose-conversion coefficients for a variety
of common radiographic projections (e.g., chest) and
views (e.g., AP, PA, lateral) for reference adult male
and female patients (Rosenstein, 1988). The anthro-
pometric characteristics of the reference patient
and the locations of the x-ray fields for each of the
projections are provided in the Handbook.
The projections (e.g., Skull) and views (e.g., AP)
tabulated are as follows.
The specifications for the projections and views
reflect practice in the United States.
The tissues or organs for which conversion
coefficients are tabulated are as follows.
The coefficients are presented for 12 beam qualit-
ies expressed in terms of HVL
1
(from 1.0 to 6.5 mm
Al in 0.5 mm Al steps) without identification of a
specific tube voltage. In preparing the tabulations,
27 diagnostic x-ray spectra were applied, 9 in each of
the categories shown below.
For the range of the peak tube voltage and con-
ventional aluminum filtration combinations used in
diagnostic radiology, the data presented should have
uncertainties of <10 % when HVL
1
alone is used to
describe beam quality (Rosenstein, 1988). Data are
also presented for the AP skull, PA chest, and AP
abdominal projections and views when an erbium
composite filter (3.2 mm erbium composite plus
2.0 mm Al) is used.
A sample tabulation for the AP thoracic spine for
the range of HVL
1
from 2.5 to 4.0 mm Al is given
in Table B.1.
Additional information is provided in the
Handbook to permit corrections for the following
circumstances.
(a) When the testes are closer to the x-ray field edge
than indicated by the reference conditions;
(b) When more or less active BM is included in
the x-ray field because of the actual field
size and;
(c) When the testes, ovaries, or uterus are shiel-
ded, particularly for the AP view in the region
of the abdomen.
Skull AP, PA, LAT Cholecystography PA
Facial bones PA Lithotripsy AP
Temporalmandibular
joint, Zygomes PA
Upper GI Tract AP,
PA, LAT
Mastoid, Mandible PA Upright abdomen AP
Temporalmandibular
joint, Sella volume LAT
Lumber spine AP, LAT
Cervical spine AP, LAT Abdominal AP, PA, LAT
Shoulder(s) AP, LAT Sacrum, Coccyx AP, LAT
Scapula(e) AP, LAT Sacroiliac joint AP
Humerus AP Pelvis, Lumbopelvic
AP, LAT
Sternum LAT Urethrogram AP
Ribs, Barium swallow
AP, PA, LAT
Cystography AP
Chest AP, PA, LAT Hip(s) AP, LAT
Thoracic spine AP, LAT Femur AP, LAT
Full Spine AP, LAT
Range of peak
tube voltage/kV
Range of HVL
1
/
mm Al
4570 1.03.6
8090 1.75.5
98120 3.16.0
Lungs
Active BM
1
Thyroid
Female breasts
Testes
Ovaries
Uterus (surrogate for the embryo; first two months
of pregnancy)
Total trunk tissue (excludes lungs and skeletal
tissue)
1
Dosimetrists term for red BM.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
70
B.2.3 Application of the handbook
The information required to estimate tissue doses
fromradiographic examinations is K
a,i
and the HVL
1
(mm Al). It is preferred that this information be
obtained for the x-ray equipment and for the phys-
ical parameters actually used at the facility.
B.2.4 Illustrative calculation
An average-sized woman who is not pregnant
undergoes a thoracic spine examination with one
AP radiograph. The examination is performed with
an HVL
1
of 2.5 mm Al, a d
FID
of 102 cm, and a field
size of 18 cm 43 cm. The dose to the female breasts
due to this examination is
D
Breast
5:3 mGy 0:411 2:18 mGy, B:1
where 5.3 mGy is the locally determined K
a,i
, and
0.411 the organ dose-conversion coefficient for the
female breast derived from Table B.1.
B.3 HANDBOOK BRH 79-8079 (ROSENSTEIN
ET AL., 1979): PROJECTIONS COMMON IN
PAEDIATRIC RADIOLOGY
B.3.1 Main features of calculations
The details of the Monte Carlo procedure used to
produce the Handbook for paediatric patients are
the same as for the Handbook for adults. The details
of the newborn, 1 year old, and 5 year old anthropo-
morphic phantoms used to produce the Handbook
are described by Hwang et al. (1976a,b). The BM of
the newborn and the 1 year old is assumed to be
100 % active BM, and that of the 5 year old reflects
the gradual replacement with yellow marrow seen
with maturation.
The variety of x-ray projections and views and
the variation in technique factors are more limited
in paediatric radiology (Beck and Rosenstein, 1979).
For the paediatric cases, each calculation started
with the specified characteristics of the x-ray
projection and view (i.e., AP, PA, LAT or OBL),
including focus-to-image receptor film distance,
d
FID
, field size and location with respect to anatom-
ical landmarks, and beam quality. X-ray spectra
were matched with the desired beam qualities
with respect to tube voltage and HVL
1
(Beck and
Rosenstein, 1979). The distribution of x-ray photon
energies was input directly at the start of the Monte
Carlo computation.
B.3.2 Scope of the handbook
The Handbook presents a series of 20 tabulations
of tissue dose-conversion coefficients for a variety of
common radiographic projections and views in pae-
diatric radiology (Rosenstein et al., 1979). The Hand-
book is for three reference patients: a newborn (birth
to 6 months), a 1 year old child and a 5 year old child.
The anthropometric characteristics of the three ref-
erence paediatric patients, the locations of the x-ray
fields for each of the projections and views, the ana-
tomical limits of the x-ray fields, and the coefficients
of variation associated with the Monte Carlo calcu-
lations are provided in the Handbook.
The projections and views tabulated are as follows.
The specifications for these projections and views
reflect practice in the United States.
Table B.1. Sample tabulation: Handbook of Tissue Doses
for Projections Common in Diagnostic Radiology (Adults)
(Rosenstein, 1988).
(D
T
/mGy)/(K
a,i
/mGy)
HVL
1
/mm Al 2.5 3.0 3.5 4.0
Male
Lungs 0.159 0.187 0.212 0.233
Active BM 0.023 0.029 0.033 0.038
Thyroid 0.107 0.123 0.137 0.148
Trunk tissue 0.067 0.078 0.087 0.095
Testes
a a a a
Female
Lungs 0.120 0.140 0.159 0.175
Active BM 0.019 0.024 0.029 0.032
Thyroid 0.107 0.123 0.137 0.148
Trunk tissue 0.054 0.062 0.070 0.075
Breasts 0.411 0.444 0.470 0.493
Ovaries
a
0.001 0.001 0.002
Uterus
a a
0.001 0.002
AP Thoracic Spine: d
FID
102 cm; field size at film, 18 cm 43 cm.
a
(D
T
/mGy)/(K
a,i
/mGy) < 0.001.
Townes
*
(all ages) Bladder AP (5 year old)
Skull AP, LAT
(all ages)
Bladder OBL
**
(1 and
5 year old)
Skull PA (1 and
5 year old)
Erect abdomen AP
(5 year old)
Neck AP, LAT
(1 year old)
Abdomen AP, PA, LAT
(all ages)
Chest AP, PA, LAT
(all ages)
Pelvis AP (all ages)
Chest OBL
**
(5 year old)
Hip AP (5 year old)
Kidneys AP
(all ages)
Hip OBL
**
(1 and
5 year old)
*
Townes field incident at 30

caudal on the anterior skin


surface.
**
OBL views field incident at 45

to the normal on the anterior


skin surface (i.e., posterior oblique).
APPENDIX B
71
The tissues or organs for which conversion coeffi-
cients are tabulated are as follows.
The coefficients are presented for three or four
beam qualities (i.e., HVL
1
of 2.0, 2.5, 3.0, and
3.5 mm Al), as appropriate for the particular projec-
tion. The beam qualities listed are nominal values
and have been selected and matched to x-ray spectra
on the basis of an observed narrow range of tube
voltage, total filtration, and waveform. In preparing
the tabulations, six diagnostic x-ray spectra were
applied, as shown below.
The selection of the paediatric projections and the
technique factors associated with each type of exam-
ination have been described by Beck and Rosenstein
(1979).
A sample tabulation for the AP abdomen projec-
tion and view for HVL
1
of 2.0 and 2.5 mm Al is given
in Table B.2.
B.3.3 Application of the handbook
The information required to estimate tissue doses
from paediatric radiographic examinations is the
K
a,i
and the HVL
1
(mm Al). It is preferred that this
information be obtained for the x-ray equipment and
for the physical parameters actually used at the
facility.
B.3.4 Illustrative calculation
A1 year old girl is to undergo a single AP abdomen
radiograph. The examination is performed with an
HVL
1
of 2.5 mm Al, a d
FID
of 102 cm, and a field size
of 18 cm 21 cm (imaging of body part). The dose to
the ovaries due to this examination is
D
Breast
0:5 mGy 0:422 0:21 mGy, B:2
where 0.5 mGy is the locally determined K
a,i
, and
0.422 the organ dose-conversion coefficient for the
ovaries derived from Table B.2.
B.4 HANDBOOK CDRH 85-8239
(ROSENSTEIN ET AL., 1985):
MAMMOGRAPHY
B.4.1 Main features of calculations
The details of the Monte Carlo procedure used to
produce the Handbook for mammography are the
same as for the Handbooks for adults and paediatric
patients. The female breast is a variable composite
of adipose and glandular tissue, and the glandular
tissue is the tissue at risk for breast cancer. The
details of the simulated compressed reference
breasts used to produce the Handbook are described
by Rosenstein et al. (1985).
Tissue-air-ratios for a plane parallel x-ray field
(20 cm 25 cm) normally incident on and fully
encompassing each reference breast were computed
for 100 % glandular and 100 % adipose compositions.
Appropriate geometrical conversions are made for
Testes Active BM
Ovaries Lungs
Thyroid Total body (includes all body tissues)
Peak tube
voltage/kV
Total filtration/
mm Al
HVL
1
/
mm Al
65 2.7 2.1
80 2.7 2.4
90 2.7 2.7
80 3.7 3.0
98 2.5 3.1
105 2.5 3.4
Table B.2. Sample tabulation: Handbook of Tissue
Doses for Projections Common in Pediatric Radiology
(Rosenstein et al., 1979).
AP abdomen Newborn 1 year old 5 year old
d
FID
/cm 102 102 102
Field size at image receptor to:
Body part/cm cm 13 13 18 21 23 30
Film size/cm cm 20 25 25 30 28 36
(D
T
/mGy)/(K
a,i
/mGy)
HVL
1
/mm
Al collimation
2.0 2.5
Organ Age Body
part
Film
size
Body
part
Film
size
Testes Newborn 0.098 1.039 0.164 1.142
1 year (0.120) (1.221) (0.120) (1.221)
5 year (0.143) (1.221) (0.143) (1.221)
Ovaries Newborn 0.445 0.445 0.639 0.639
1 year 0.308 0.308 0.422 0.422
5 year 0.308 0.308 0.422 0.422
Thyroid Newborn (0.006) (0.029) (0.006) (0.029)
1 year
a
(0.010)
a
(0.010)
5 year
a
(0.003)
a
(0.003)
Active BM Newborn 0.104 0.182 0.145 0.241
1 year 0.079 0.113 0.114 0.160
5 year 0.063 0.079 0.095 0.115
Lungs Newborn 0.056 0.501 0.075 0.567
1 year 0.040 0.259 0.055 0.291
5 year 0.045 0.116 0.054 0.140
Total body Newborn 0.212 0.428 0.242 0.478
1 year 0.191 0.300 0.224 0.348
5 year 0.180 0.231 0.215 0.273
Values in parenthesis indicate differences in (D
T
/mGy)/(K
a,i
/mGy)
as a function of HVL
1
for the same collimation, i.e., body part or
film size, were not significant.
a
(D
T
/mGy)/(K
a,i
/mGy) < 0.001.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
72
the divergent x-ray field to obtain the tissue dose per
unit incident air kerma. The computations were for
monoenergetic photons with seven selected energies
from 9.5 to 55.5 keV, with additional data interpol-
ated at 0.5 keV intervals. The coefficients of vari-
ation associated with the tissue-air-ratios for the
seven selected energies are provided in Anderson
and Rosenstein (1985). The tissue-air-ratio for any
intervening breast composition is derived from the
relevant mass fractions of glandular and adipose
tissue.
For a given breast composition, the tissue-
air-ratios for monoenergetic photons are then con-
verted to tissue-air-ratios for a number of x-ray spec-
tra that represent the tungsten, tungsten
molybdenum alloy and molybdenum target materi-
als then used in mammography techniques (Fewell
and Shuping, 1978). Appropriate geometrical con-
versions are made for the divergent x-ray field to
obtain the tissue dose per unit incident air kerma.
The glandular tissue dose per unit incident air
kerma is obtained by multiplying the total breast
tissue dose by the ratio of the mass energy absorp-
tion coefficient of glandular tissue to that of the
selected overall breast composition. The calculations
are described by Rosenstein (1984) and Andersen
and Rosenstein (1985).
B.4.2 Scope of the handbook
The Handbook presents tabulations of glandular
tissue dose-conversion coefficients resulting from
the Monte Carlo calculations for five reference
breasts and adds one tabulation that reproduces
the experimental results of Stanton et al. (1984).
The Handbook permits the user to compute glandu-
lar tissue dose for various techniques in mammo-
graphy as a function of breast compression, breast
size, breast composition, and the quality of the x-ray
beam (Rosenstein et al., 1985).
The six tabulations are listed below; CC stands
for cranio-caudal and ML stands for medio-lateral.
The specifications for these tabulations reflect
practice in the United States.
The five tabulations resulting from the Monte
Carlo calculations are presented as a function of
HVL
1
(from 0.2 to 2.4 mm Al) without identification
of a specific tube voltage. In preparing the tabula-
tions, 51 mammographic x-ray spectra were used, as
listed below.
For the same target type and beam quality, but
different tube voltages, the individual values were
within 10 % of the average over the range of tube
voltages. For similar HVL
1
using different target
materials, the individual values are within a few
percent of the average for the three target types
(Rosenstein, 1984). The values in the tabulations
result from smooth curves fitted to the results from
the individual x-ray spectra.
The tabulation that reproduces the experimental
results of Stanton et al. (1984) includes values for
tungsten targets and separate values for molyb-
denum or tungstenmolybdenum targets.
The portion of the tabulation for the CC view, 6 cm
uniform breast thickness, glandular tissue content
between 25 and 75 %, and HVL
1
from 0.2 to 1.0 mm
Al, is given in Table B.3.
B.4.3 Application of the handbook
The information required to estimate glandular
tissue dose from mammography examinations is
K
a,i
and the HVL
1
(mm Al). It is preferred that this
information be obtained for the mammography
equipment and for the physical parameters actually
used at the facility.
B.4.4 Illustrative calculation
A women undergoes a single CC view mammo-
gram. The breast is compressed to 6 cm and the
composition is 25 % glandular tissue. The exami-
nation is performed with an HVL
1
of 0.6 mm Al,
a d
FID
of 76 cm, and a field size of 20 cm 25 cm.
The mean dose to the glandular tissue due to this
examination is
D
Breast
6:1 mGy 0:292 1:78 mGy, B:3
Firm compression (uniform breast thickness)
1. CC view: 3 cm to 8 cm uniform thickness;
50 % glandular tissue (Stanton et al., 1984)
2. CC view: 6 cm uniform thickness;
5100 % glandular tissue
Moderate compression (non-uniform breast
thickness)
3. CC view: small breast; glandular tissue 5100 %
4. CC view: medium breast; glandular tissue 5100%
5. CC view: large breast; glandular tissue 5100 %
6. ML view: medium breast; glandular
tissue 5100%
Target
type
No. of
spectra
Range of
tube
voltage/kV
Range of
HVL
1
/mm
Al
Tungsten (W) 19 2550 0.42.3
MoW alloy 14 2050 0.21.6
Molybdenum (Mo) 18 3050 0.31.3
APPENDIX B
73
where 6.1 mGy is the locally determined K
a,i
, and
0.292 the organ dose-conversion coefficient for the
ovaries derived from Table B.3.
B.5 HANDBOOK CDRH 92-8282 (ROSENSTEIN
ET AL., 1992): UPPER GASTROINTESTINAL
(UPPER GI) FLUOROSCOPIC EXAMINATION
B.5.1 Main features of calculations
The Monte Carlo procedure and the anthropo-
morphic phantoms (ADAM and EVA) used to pro-
duce the Handbook for Upper GI examinations is
the adaptation of the BRHGAM code (Warner, 1973)
by the GSF Forschungszentrum fu r Umwelt und
Gesundheit (Kramer et al., 1982). The mathematical
phantom representing the reference male patient
(ADAM) is a modification of the original MIRD-5
phantom (Snyder et al., 1969), including an update
of the active BM distribution by Cristy (1981). The
reference female patient (EVA) is the ADAM phan-
tom reduced uniformly to 83 % of its original size,
with the unique female tissues instead of the unique
male tissues (Kramer et al., 1982).
To better represent the upper GI simulations
an esophagus and a duodenum were added to the
phantom, and a double contrast medium was simu-
lated in the esophagus, duodenum, and stomach
(Stern et al., 1995a). The contrast medium consisted
of a 1 mm thick lining of barium sulphatewater
against the inner organ walls, with air filling the
rest of the volume of the organ.
The dynamic upper GI fluoroscopic examination
was approximated with a set of six discrete x-ray
fields developed by Suleiman (1989). Each discrete
x-ray field is a projection of distinct anatomical por-
tions of the upper GI tract: upper, middle, or lower
esophagus, gastroesophageal junction, stomach, and
duodenum. For each component projection,
a separate Monte Carlo calculation was made with
ADAM and EVA phantoms for the relevant oblique
x-ray fields in divergent-beam geometry, with spe-
cific x-ray spectra representative of clinical practice.
B.5.2 Scope of the handbook
The Handbook presents a series of 12 tabulations
of tissue dose-conversion coefficients, one tabulation
for each of two views i.e., left posterior oblique (LPO)
and right anterior oblique (RAO) for each of the six
component anatomical projections of the Upper GI
fluoroscopic examination (Rosenstein et al., 1992).
Descriptions of the anatomical projections, the
oblique views, the locations of the projections on
the reference phantoms, the simulation of the con-
trast medium, and the coefficients of variation for
the Monte Carlo calculations are provided in the
Handbook and by Stern et al. (1995a).
The anatomical projections (e.g., Upper esopha-
gus) and views (e.g., LPO) tabulated are as follows.
The specifications for these projections and views
reflect practice in the United States.
The tissues or organs for which conversion coeffi-
cients are tabulated are as follows.
The coefficients are presented for three beam
qualities (HVL
1
of 4.0, 5.0, and 5.5 mm Al). These
beam qualities span the range of fluoroscopic x-ray
tube potentials and aluminum filtrations used in
clinical practice (Conway, 1994). The beam qualities
cited approximate those measured above a tabletop
present in the beam path of undertable x-ray tubes.
For the usual range of tube voltages and aluminum
filtration combinations used in Upper GI fluoro-
scopy, the results should have uncertainties of
<10 % when HVL
1
alone is used to describe beam
Table B.3. Sample tabulation: Handbook of Glandular
Tissue Doses in Mammography (Rosenstein et al., 1985).
(D
T
/mGy)/(K
a,i
/mGy)
Glandular tissue content (by weight) 25% 50% 75%
HVL
1
/mm Al
0.2 0.070 0.061 0.054
0.3 0.137 0.121 0.110
0.4 0.190 0.170 0.154
0.5 0.242 0.218 0.201
0.6 0.292 0.266 0.247
0.8 0.384 0.354 0.332
1.0 0.458 0.427 0.404
Craniocaudal view: Firm compression (uniform breast thickness),
thickness 6 cm.
Upper esophagus (LPO and RAO)
Middle esophagus (LPO and RAO)
Lower esophagus (LPO and RAO)
Gastroesophageal junction (LPO and RAO)
Stomach (LPO and RAO)
Duodenum (LPO and RAO)
Thyroid Bladder
Oesophagus Liver
Female breasts Testes
Lungs Ovaries
Active BM Uterus (surrogate for the embryo;
first 2 months of pregnancy)
Stomach Total trunk tissue (excludes lungs
and skeletal tissue)
Colon
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
74
quality. In preparing the tabulations, three dia-
gnostic x-ray spectra were applied, as shown below.
For the tabulated data, the d
FID
is 80 cm and the
d
FSD
is 50 cm. For the differences from these refer-
ence d
FID
and d
FSD
common in Upper GI fluoroscopy
(i.e., d
FSD
between 40 and 55 cmand d
FID
between 70
and 90 cm), differences from the tabulated entries
will be <20 % for entries of (D
T
/mGy)/(K
a
/mGy) >0.1
and <35 % for entries of (D
T
/mGy)/(K
a
/mGy) <0.1
(Rosenstein et al., 1992).
A sample tabulation for the stomach (RAO) is
given in Table B.4.
B.5.3 Application of the handbook
An upper GI fluoroscopic examination needs to be
analysed and simulated with the 12 specific x-ray
fields provided in the Handbook. A practical method
of accomplishing this task is to videotape a clinically
representative examination and add a permanent
visual display of time and the sequence of frames.
The videotape can then be analysed and the clinical
examination subdivided into the fluoroscopic scan
segments and radiographic spot films corresponding
to the anatomical projections and views in the Hand-
book. The irradiation time associated with each of
the fluoroscopic scansegments canbe evaluated from
the visual display. An example of this segmentation
is provided in the Handbook.
An example of the determination of K
a,i
for each
of the discrete x-ray fields and the determination of
a separate representative beam quality for fluoro-
scopic scans and radiographic spot films is also
provided in the Handbook.
B.5.4 Illustrative calculation
A male patient undergoing an illustrative
examination of the stomach is presented in the
Handbook. The examination consists of various
parts, for instance, gastroesophageal junction (GE)
projection in LPO and RAO views for both scans and
films, stomach (ST) projection in LPO and RAO
views for both scans and films, and duodenum (DU)
projection in LPO view for both scans and films.
Here only ST projection in RAO for both scans and
films are shown. The example is developed fully in
the Handbook.
The ST RAO scans are performed with an HVL
1
of 5.5 mm Al, a d
FSD
of 50 cm, a d
FID
of 80 cm, and a
field size of 22.9 cm 22.9 cm at image receptor
plane. The dose to the stomach due to this part of
the examination is
D
Stomach
4:5 mGy 0:354 1:59 mGy, B:4
where 4.5 mGy is the locally determined K
a,i
, and
0.354 the organ dose-conversion coefficient for the
stomach taken from Table B.4.
The ST RAO films are performed with an HVL
1
of 5.0 mm Al, a d
FSD
of 50 cm, a d
FID
of 80 cm,
and a field size of 22.9 cm 22.9 cm at image recep-
tor. The dose to the stomach due to this part of the
examination is
D
Stomach
17:1 mGy 0:342 5:85 mGy, B:5
where 17.1 mGy is the locally determined K
a,i
, and
0.342 the organ dose-conversion coefficient for the
stomach taken from Table B.4.
The total dose to the stomach due to ST RAO scans
and films is
D
Stomach
1:59 mGy 5:85 mGy 7:44 mGy:
Peak tube voltage/kV HVL
1
/mm Al
80 4.0
100 5.0
120 5.5
Table B.4. Sample tabulation: Handbook of Tissue Doses
for the Upper Gastrointestinal Fluoroscopic Examination
(Rosenstein et al., 1992).
(D
T
/mGy)/(K
a,i
/mGy)
HVL
1
/mm Al 4.0 5.0 5.5
Tissue Male Female Male Female Male Female
Thyroid
a a a a a a
Oesophagus 0.011 0.016 0.017 0.021 0.018 0.023
Breast 0.008 0.010 0.011
Lung 0.011 0.016 0.015 0.019 0.017 0.021
Active BM 0.014 0.013 0.017 0.015 0.018 0.016
Stomach 0.297 0.354 0.342 0.400 0.354 0.422
Colon 0.016 0.021 0.021 0.024 0.024 0.027
Bladder 0.001
a
0.001 0.001 0.001
a
Liver 0.091 0.071 0.105 0.080 0.110 0.084
Testis
a a a
Uterus 0.001 0.002 0.002
Ovary 0.002 0.003 0.002
Trunk 0.040 0.040 0.046 0.046 0.048 0.048
Stomach, RAO: The patient is usually prone with the right
anterior side against the table. The patients sagittal planes are
angled at a nominal 30

from the central ray of the x-ray beam. A


distinguishing characteristic of this projection is that it is centred
on the stomach.
Field size at image receptor: 22.9 cm 22.9 cm
d
FID
: 80 cm
d
FSD
: 50 cm
Note: Entries apply to both single and double contrast procedures.
a
(D
T
/mGy)/(K
a,i
/mGy) < 0.001.
APPENDIX B
75
B.6 HANDBOOK CDRH 95-8289 (STERN
ET AL., 1995B): FLUOROSCOPIC AND
CINEANGIOGRAPHIC EXAMINATION OF
THE CORONARY ARTERIES
B.6.1 Main features of calculations
The Monte Carlo procedure and the anthropo-
morphic phantoms (ADAM and EVA) used to pro-
duce the Handbook for examinations of the coronary
arteries is the adaptation of the BRHGAM code
(Warner, 1973) by the GSF Forschungszentrum
fu r Umwelt und Gesundheit (Kramer et al., 1982).
The mathematical phantom representing the refer-
ence male patient (ADAM) is a modification of the
original MIRD-5 phantom (Snyder et al., 1969), with
the addition of an esophagus by Zankl et al. (1992)
and the addition of a supporting tabletop at the back
of the phantoms (Stern et al., 1995b). The reference
female patient (EVA) is the ADAMphantom reduced
uniformly to 83 %of its original size, with the unique
female tissues instead of the unique male tissues
(Kramer et al., 1982).
The simulated examinations of the coronary arter-
ies are based on a series of distinct x-ray fields com-
monly used in coronary interventional radiology,
derived from analyses of practice at the Institut de
Cardiologie de Montreal (Lesperance, 1982; Haddadi
and Renaud, 1993). The scope of views and arterial
projections represented are generally applicable to a
broad range of examinations following a variety of
clinical protocols in many different venues. For each
arterial projection, a separate Monte Carlo calcula-
tion was made with ADAM and EVA for the relevant
complex oblique x-ray field in divergent-beam geo-
metry, with specific x-ray spectra representative of
clinical practice (Stern et al., 1995b).
B.6.2 Scope of the handbook
The Handbook presents 12 tabulations of tissue
dose-conversion coefficients, 1 tabulation for each
of 11 arterial projections, and 1 tabulation that
provides nominal data that can be used for an exam-
ination consisting of several views, in lieu of the
individual tabulations. The nominal approach is
acceptable for coronary-artery examinations
because all the views share the point that the heart
is a relatively small, common, central region inter-
cepted by the central ray of each different x-ray field.
Descriptions of the standard nomenclature used in
the anatomy of the coronary arteries and to identify
the arterial projections, the specifications for the
irradiation geometry and complex oblique views,
and the coefficients of variation for the Monte Carlo
calculations are provided in the Handbook (Stern
et al., 1995b).
The arterial projections are specified by view (e.g.,
RAO), angulation of image receptor (if different
from 0

) in the transverse and saggital planes, and


location of the field centre (left or right ventricle).
The 12 tabulations are as follows.
The tissues or organs for which conversion coeffi-
cients are tabulated are as follows.
In extended coronary-artery examinations, cumu-
lative absorbed doses to that portion of skin lying
directly in the path of the incident primary field may
approach or exceed the thresholds for deterministic
injury. Therefore, the SI units for the conversion
coefficients in this Handbook are expressed as
(D
T
/mGy)/(K
a,i
/Gy). The entrance skin in the prim-
ary field is only a small fraction of the entire skin
tissue; the extent is delimited and the location is
determined by the collimation and irradiation geo-
metry of the anatomical projections.
Except for the heart, the organs and tissues listed
are those with which cancer, genetic effects, or
in utero effects have been associated with irradiation
(ICRP, 1991b). The heart surrounds the ventricle
field centres and always lies within the field of view
(FOV). It receives the highest absorbed dose per unit
of incident air kerma of the internal organs. The
data for the heart are provided for reference pur-
poses; there is no health effect yet established for
absorbed dose in the heart for the ranges that occur
in coronary-artery examinations.
RAO 30

(left) Cranial 20

(right)
RAO 30

(right) LAO 45

Cranial 25

(left)
LAO 30

(right) LAO 45

(left)
LLAT (left) RAO 10

Cranial 40

(left)
RAO 15

Caudal 25

(left)
Anterior (left)
RAO 15

Cranial 25

(left)
Nominal conversion
coefficients
Entrance skin in primary field
Brain Pancreas
Thyroid Stomach
Thymus Liver
Active BM Kidneys
Oesophagus Colon
Lungs Small intestine
Female breasts Ovaries
Heart Uterus (surrogate for the embryo)
Adrenals Urinary bladder
Spleen Testes
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
76
The coefficients are presented for three beam
qualities for each reference patient (i.e., ADAM,
HVL
1
of 2.5, 4.0, and 5.5 mm Al; EVA, HVL
1
of 2.0,
3.5, and 5.0 mm Al). The ranges of beam qualities
correspond to the those observed in the study con-
ducted at the Institut de Cardiologie de Montreal
(Haddadi and Renaud, 1993) and that observed in a
nationwide survey of fluoroscopy practice in the Uni-
ted States (Conway, 1994). The beam qualities are
without the presence of a supporting tabletop. For
the usual ranges of tube voltage (kV) and aluminum
filtration combinations used in fluoroscopy and cine-
angiography of the coronary arteries, the results
should have uncertainties of <10 % when HVL
1
alone is used to describe beam quality. In preparing
the tabulations, the following diagnostic x-ray spec-
tra were applied, as shown below.
When the actual diameter of the FOV at the image
receptor plane differs fromthat used in the reference
simulation by >20 %, the dose-conversion factor in
Table B.5 should be corrected by multiplication with
[FOV(actual)/FOV(tabulated)]
2
. The basis for the
correction is given in the Handbook (Stern et al.,
1995b).
The correction is not applicable to the entrance
skin in the primary field, because the absorbed dose
does not depend on the size of the area of the skin
irradiated.
A tabulation for the conversion coefficients is
given in Table B.5.
B.6.3 Application of the handbook
The Handbook can be used to perform a view-
by-view or a nominal analysis of an examination.
To use the nominal approach, the coronary-artery
examination is characterized in an overall sense.
This permits a quick, but somewhat less accurate
way to estimate nominal tissue doses for a complete
examination without detailed specifications for the
particular views applied clinically.
The complete examination is characterized with
nominal values for four parameters.
(a) The HVL
1
;
(b) The total K
a,i
for all the fluoroscopic plus cine-
angiographic segments (i.e., summed for all
skin-entrance planes, wherever these planes
are located);
(c) The FOV diameter at the image receptor plane;
and
(d) The highest cumulative K
a,i
(i.e., fluoroscopic
plus cineangiographic) for any single skin
location. Such a skin region may be irradiated
in only one view or possibly in multiple views
that share a common locus of irradiation.
B.6.4 Illustrative calculation
The illustration below is for a left-heart study of
an adult male entailing a left ventriculogram in
biplanar mode followed by left and right coronary
Peak tube
voltage/kV
Total filtration/
mm Al
HVL
1
/mm Al
ADAM
60 3.5 2.5
90 4.0 4.0
120 4.3 5.5
EVA
50 3.3 2.0
80 3.9 3.5
110 4.2 5.0
Table B.5. Sample tabulation: Handbook of Tissue Doses
for Fluoroscopic and Cineangiographic Examinations of
the Coronary Arteries (Stern et al., 1995b).
(D
T
/mGy)/(K
a,i
/Gy)
Male Female
HVL
1
/mm Al 2.5 4.0 5.5 2.0 3.5 5.0
Entrance skin
in primary field
1000 1120 1180 950 1090 1170
Brain 0.003 0.020 0.041 0.001 0.018 0.045
Thyroid 0.12 0.50 0.85 0.075 0.53 1.1
Thymus 2.2 6.5 9.9 1.6 6.7 12
Active BM 6.1 12 16 5.2 12 17
Oesophagus 14 33 47 11 32 51
Lung 34 53 65 31 55 71
Breast 3.0 9.4 15
Heart 30 62 86 23 63 95
Adrenal 36 62 78 32 64 87
Spleen 4.9 11 15 3.8 11 17
Pancreas 5.4 14 20 3.8 13 22
Stomach 2.4 6.3 9.3 1.7 6.3 10
Liver 4.4 9.7 14 3.5 9.9 15
Kidney 3.2 6.8 9.3 2.7 7.1 11
Colon 0.071 0.31 0.56 0.043 0.32 0.67
Small intestine 0.091 0.40 0.72 0.050 0.39 0.82
Ovary 0.007 0.076 0.19
Uterus 0.005 0.071 0.17
Urinary bladder 0.003 0.021 0.044 0.001 0.023 0.054
Testis
a
0.002 0.004
Nominal conversion coefficients:
d
FSD
: 60 cm
d
FID
: 90 cm
FOV diameter at image receptor: 14 cm.
a
(D
T
/mGy)/(K
a,i
/Gy) < 0.001.
APPENDIX B
77
angiography. In a typical application of the nominal
approach, the user will rely on estimated values of
the parameters for the entire examination that may
be truly nominal. In any case the estimates will be
developed at the level of detail and with the degree
of accuracy available to the user.
The left-heart study is performed with an HVL
1
of 3.6 mm Al, a d
FSD
of 60 cm, a d
FID
of 90 cm, and a
FOV of 20 cm at image receptor plane.
The dose to the skin in primary field due to the
examination is
D
Skin;local
0:80 Gy 1090 mGy=Gy 870 mGy,
B:6
where 0.80 Gy is the highest K
a,i
for any single skin
location and 1090 the organ dose-conversion coeffi-
cient for the local skin derived by linear interpola-
tion between the reference values for HVL
1
from
Table B.5.
The dose to the active BM due to the examination is
D
BM
1:6 Gy 10 mGy=Gy 2:0 32 mGy,
B:7
where 1.6 Gy is the K
a,i
for all skin entrances planes,
10 the organ dose-conversion coefficient for the act-
ive BM derived by linear interpolation between the
reference values for HVL
1
from Table B.5, and 2.0 is
a correction factor needed for the actual FOV (20 cm
rather than the 14 cm in the reference tabulation in
Table B.5).
The dose to the lung due to the examination is
D
BM
1:6 Gy 48 mGy=Gy 2:0 150 mGy,B:8
where 1.6 Gy is the K
a,i
for all skin entrances planes,
48 the organ dose-conversion coefficient for the lung
derived by linear interpolation between the refer-
ence values for HVL
1
from Table B.5, and 2.0 is a
correction factor needed for the actual FOV.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
78
Journal of the ICRU Vol 5 No 2 (2005) Report 74 doi:10.1093/jicru/ndi031
Oxford University Press
APPENDIX C: REPORTS PRODUCED BY THE
GERMAN NATIONAL RESEARCH CENTER FOR
ENVIRONMENT AND HEALTH (GSF)
C.1 GENERAL DESCRIPTION
The GSF, Institute for Radiation Protection,
Medical Physics Department has produced several
reports containing tabulations of organ dose-
conversion coefficients for reference adult and pae-
diatric patients for common radiographic and CT
examinations. The reports permit the calculation of
organ doses from measurable normalization quant-
ities for specified technical parameters relating to
the examinations, using the general relation:
Organ dose organ dose-conversion coefficient
normalization quantity: C:1
For standardized technical parameters, as given
in the European Guidelines on Quality Criteria for
Diagnostic Radiographic Images (EC, 1996a), the
organ doses are given.
The following material is available.
For the calculations on adult patients, gender-
specific mathematical human phantoms have been
used, whereas the calculations in paediatric radi-
ology have used voxel phantoms. The adaptation
of the data to individual patients is discussed.
Details are described in the GSF Report Series:
The Calculation of Dose from External Exposures
Using Reference Human Phantoms and Monte
Carlo Methods (Kramer et al., 1982; Drexler et al.,
1990; 1993; Zankl et al., 1991; 1993; 1997).
C.2 PROJECTIONS COMMON IN
DIAGNOSTIC RADIOLOGY (ADULTS)
GSF-BERICHT 11/90 (DREXLER
ET AL., 1990)
C.2.1 Main features of calculations
The organ dose-conversion coefficients are
obtained by applying a Monte Carlo calculation
method (Kramer et al., 1982; Veit et al., 1989) to
the modified MIRD-5 type mathematical male
ADAM and female EVA models (Kramer et al.,
1982). The main differences of the gender-specific
ADAM and EVA phantoms in comparison to the
MIRD-5 type phantoms are:
(i) the EVA phantom is representative of a refer-
ence adult female and contains the female
organs,
(ii) the ADAM phantom is representative of a
reference adult male and contains the male
organs,
(iii) the EVA phantom is significantly smaller than
the ADAM phantom.
(iv) the arms of both phantoms are removable,
(v) the phantoms have a chin so that less tissue
exists in front of the thyroid,
(vi) the skull of ADAM and EVA is improved with
regard to shape and size, and
(vii) an esophagus has been added (Zankl et al.,
1992).
In the Monte Carlo calculations the beam quality
is described by the spectral distributions of the
photons, which is a function of the tube voltage, the
total filtration, the target material, and the target
angle as determined by a computer program based
on semi-empirical methods (Birch and Marshall,
1979). In the tabulations of organ dose-conversion
coefficients, the beam quality is expressed in terms
of the constant tube voltage and the total filtration.
1. Projections Common
in Diagnostic Radiology
(Adults)
(GSF-Bericht 11/90,
Drexler et al., 1990).
2. Projections Common
in Diagnostic Radiology
(Paediatric)
(Zankl et al., 1988;
1989).
3. Organ Doses for
CT Examinations
of Adults
(GSF-Bericht 30/91,
Zankl et al., 1991).
4. Organ Doses for
CT Examinations in
Pediatric Radiology
(GSF-Bericht 30/93,
Zankl et al., 1993).
International Commission on Radiation Units and Measurements 2005
The beam geometry is given by the focus-to-
surface distance d
FSD
, the focus-to-image receptor
distance d
FID
, the field size at the image receptor
plane, and the direction of incidence on the human
body. The organ dose-conversion coefficients for each
projection are calculated by simulating 2 10
6
incid-
ent photons for the exposure under consideration.
The exposure conditions have been selected to be in
close accordance with routine exposure conditions
common in Europe.
Details are described in the GSF Report Series:
The Calculation of Dose from External Exposures
Using Reference Human Phantoms and Monte Carlo
Methods (Kramer et al., 1982; Drexler et al., 1990;
1993).
C.2.2 Scope of the report
The Report of Drexler et al. (1990) presents a ser-
ies of tabulations of organ dose-conversion coeffi-
cients for 40 common radiographic examinations
including techniques for pregnant women
(Table C.1). Additional organ dose tabulations are
provided for 28 projections with technical paramet-
ers as published in the EC Quality Criteria Docu-
ment (1996a). The organs and tissues for which
organ dose-conversion coefficients are presented is
listed in Table C.2.
A sample tabulation for the lung PA projection is
given in Table C.3. Organ dose-conversion coeffi-
cients are provided for the ADAM and the EVA
phantoms. Each x-ray examination is investigated
for three tube voltages. Atotal filtration of 2.5 mmAl
is used for all the radiographic projections. The last
column in the table presents the coefficient of vari-
ation of the organ dose-conversion coefficients estim-
ated from the Monte Carlo calculation.
C.2.3 Application of the report
Organ doses D
T
are determined from incident
air kerma K
a,i
using the organ dose-conversion
coefficients c
T,K
a,i
given in the Report as,
D
T
c
T,K
a,i
K
a,i
: C:2
The value of K
a,i
for the relevant radiographic pro-
jection and the applied tube voltage, field size, and
filtration are required to estimate the D
T
. For cases
where K
a,i
is not known, the report presents a graph
to estimate K
a,i
from tube voltage, filtration, tube-
current exposure-time product, P
It
, and the focal
Table C.1. Radiographic examination of adults included
in Drexler et al. (1990).
Skull
a
AP, PA, LAT Sacrum AP, LAT
Sinuses PA Pelvis AP
Shoulder joint AP Abdomen AP
Cervical spine
a
AP, PA, LAT Kidneys AP
Thoracic spine AP, LAT Gall bladder PA
Ribs AP Bladder AP
Lung
a
AP, PA, left LAT,
right LAT
Colon contrast enema
PA, LAT
Stomach PA Right hip joint AP
Duodenum PA Left femur AP
Lumbar spine sacrum AP, LAT
a
At various d
FSD
.
Table C.2. Tissues or organs for which organ dose-
conversion coefficients are given for radiographic
examinations of adults.
Brain Colon upper
Eye lenses Colon lower
Thyroid Ovaries
Breast Uterus
Lungs Testes
Spleen Active bone marrow
Pancreas Skeleton
Stomach wall Skin entrance
Small intestine Skin exit
The German edition (Drexler et al., 1993) contains in addition
data for the following.
Oesophagus Bladder wall
Adrenals Muscle
Liver Skin
Kidneys
Table C.3. Sample table from Drexler et al. (1990) showing
mean organ dose-conversion coefficients, i.e., mean organ
dose per unit K
a,i
for the lung PA projection for adult
females (f ) and males (m).
Field size (f ) 35 cm 35 cm d
FID
180 cm
Field size (m) 35 cm 40 cm d
FSD
150 cm
Patient thickness (f ) 19 cm Total filtration 2.5 mm Al
Patient thickness (m) 20 cm
Tube voltage/kV 90 125 140 c.v
a
Organ m f m f m f
Thyroid 0.06 0.06 0.10 0.10 0.12 0.12 0.035
Breast 0.13 0.19 0.22 0.006
Lungs 0.54 0.57 0.68 0.72 0.73 0.76 0.002
Spleen 0.25 0.18 0.33 0.24 0.35 0.26 0.009
Pancreas 0.17 0.12 0.25 0.17 0.27 0.19 0.012
Stomach wall 0.07 0.06 0.11 0.10 0.12 0.11 0.013
Small intestine d d d d 0.01 0.01 0.02
Colon (upper)
b
d d 0.01 0.01 0.01 0.01 0.035
Colon (lower)
c
d d d d d d
Active bone marrow 0.11 0.12 0.16 0.17 0.17 0.18 0.001
Skeleton 0.30 0.32 0.35 0.38 0.36 0.39 0.001
Surface (entrance) 1.24 1.24 1.27 1.32 1.35 1.34 0.001
Surface (exit) 0.04 0.05 0.07 0.08 0.09 0.10 0.04
a
Coefficient of variation.
b
Caecum Ascending Transverse.
c
Descending Sigmoid Rectum.
d
Less than 0.01.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
80
spot-to-surface distance, d
FSD
. In addition, recom-
mendation is given in the Report on how to proceed,
when physical exposure parameters in a special
case considered differ from those assumed in the
tabulations.
C.2.4 Illustrative calculation
Breast dose from a PA chest examination per-
formed with a tube voltage of 125 kV, 2.5 mm Al
filtration, a P
It
of 4 mAs, a d
FSD
of 150 cm, corres-
ponding to a d
FID
of 180 cm and a field size of 35 cm
35 cm at d
FID
for the female phantom
D
Breast
0:19 0:4 mGy 0:076 mGy, C:3
where 0.19 is the organ dose-conversion coefficient
for breast fromTable C.3 and 0.4 mGy the K
a,i
, either
known from measurement or derived from the graph
in GSF-Bericht 11/90 (Drexler et al., 1990).
C.3 COMMON PROJECTIONS IN
DIAGNOSTIC RADIOLOGY (PEDIATRIC)
(ZANKL ET AL., 1988; 1989)
C.3.1 Main features of calculations
The features of the Monte Carlo program
employed are the same as that used for the adults.
The phantoms used are tomographic models of two
real patients, i.e., an 8-week-old girl BABY and a 7-
year-old girl CHILD. Details of these phantoms are
presented by Zankl et al. (1988) and Veit et al.
(1989).
To construct the GSF paediatric phantoms, whole-
body CT data were used. For each phantom, scans of
a single patient were used, performed as contiguous
slices of 4 and 8 mm thickness for the BABY and
CHILD, respectively. The baby was 8 weeks old
when it was scanned and thus represents babies
from newborn up to several months. The child was
7 years old; it was, however, slightly small for its
age and can be regarded as representative for the
ages between 5 and 7 years. The CT-based voxel
phantoms are more realistic for dose calculations
than any other type of phantom available thus far.
They do not represent an average over all persons of
a certain age and gender, as the shape of the body
and the shape and locations of the organs vary for
people of the same age and gender, and even for
different attitudes (standing, sitting, lying).
However, because the shape and location of the
organs is accurate for a particular person in a certain
attitude, the phantoms constructed from these data
give the best approach to reality. For organ dose
calculations, the body is modelled using seven
types of tissue. The distribution of active BM in
bone is derived from the CT-numbers belonging to
each voxel, thus providing realistic modelling of
this tissue.
C.3.2 Scope of the publications
Organ doses for BABY and organ dose-conversion
coefficients for CHILD for the most common types of
examination in paediatric radiology have been pub-
lished (Zankl et al., 1988; 1989) and are listed in
Tables C.5 (BABY) and C.7 (CHILD). The physical
exposure conditions for which they were determined
are shown in Tables C.4 (BABY) and C.6 (CHILD).
For BABY the tube voltage, the field sizes, and the
projections were chosen to be representative of
patients between the ages of 4 and 10 weeks who
had been x-rayed in the last 10 years in Munichs
University Children Hospital (Zankl et al., 1988).
For the thorax and pelvis examinations no grid was
used and the film was assumed to be in contact with
the body. For the other examinations grids were
used and the film was at 3.0 cm behind the surface.
Organ dose scaling factors enable the derivation of
organ doses for children deviating in size from the
original phantoms (Veit and Zankl, 1992, 1993).
The exposure conditions shown in Table C.6 for
CHILD are typical for the practice in Germany for
examinations of 57-year-old children with respect
to field size, tube voltage, filtration, focus-to-film
distance, and patient-to-film distance (Zankl et al.,
1989).
Table C.4. Exposure parameters used as input data for the dose calculation of the BABY phantom.
Type of examination Voltage/kV Height of field/cm Width of field/cm d
FID
a
/cm P
KA
a
/mGy cm
2
K
a,i
a
/mGy
Thorax AP 65 13.8 14.5 150 6.93 39.7
Abdomen AP 70 20.0 16.3 100 19.4 76.2
Pelvis AP 70 11.6 13.9 100 7.9 56.5
Skull AP 70 22.9 15.1 100 69.8 275
Skull LAT 65 19.8 20.1 100 51.3 179
a
d
FID
, focus-to-image receptor distance; P
KA
, dose-area product; K
a,i
, incident air kerma.
APPENDIX C
81
C.3.3 Application of the publications
For BABY, organ doses are presented for specific
exposure conditions and can be directly taken from
Table C.5. Organ doses for values of K
a,i
different
from that listed in Table C.4 can be easily
derived by normalizing them to the values given in
the table.
For CHILD, the quantity required to apply the
organ dose-conversion coefficients c
T,K
a,i
from
Table C.5. Organ doses D
T
/mGy for an 8-week-old BABY phantom for different x-ray procedures (as listed in Table C.4).
Type of examination Thorax AP Abdomen AP Pelvis AP Skull AP Skull LAT
Body organs and tissues D
T
/mGy D
T
/mGy D
T
/mGy D
T
/mGy D
T
/mGy
Bladder wall 0.02 60.9 47.6 0.02 0.00
Brain 0.28 0.04 0.00 88.9 60.2
Eye lenses 0.86 0.00 0.00 295 125
Heart 23.3 37.1 0.14 6.99 1.66
Lungs 18.6 23.8 0.11 16.6 2.68
Ovaries 0.02 52.1 42.1 0.03 0.00
Small intestine wall 0.13 69.0 25.8 0.26 0.07
Stomach wall 0.88 70.0 0.99 0.78 0.29
Testes 0.00 10.1 67.3 0.01 0.03
Thymus 27.4 4.38 0.06 51.7 6.04
Thyroid 34.2 1.12 0.02 212 90.3
Total skeleton 22.1 39.1 15.7 304 186
Active BM pelvis 0.01 29.2 22.1 0.07 0.04
Active BM ribs 7.53 13.3 0.10 11.7 1.46
Active BM skull 1.43 0.07 0.00 86.9 52.1
Total active BM 3.02 5.83 2.81 43.9 22.8
Table C.6. Exposure parameters used as input data for the dose calculation of the CHILD phantom, relevant for 57-year-
old patients in Germany.
Type of examination Voltage/kV Field width/cm Field height/cm d
FID
a
/cm d
PID
a
/cm
Skull AP 70 26.5 31.0 110 10
Skull LAT 70 26.5 30.5 110 10
Thorax AP 60 25.7 19.0 100 0
Thorax PA 60 26.7 19.7 150 10
Abdomen AP 70 25.6 39.4 110 10
Pelvis AP 70 26.2 22.7 110 10
a
d
FID
focus-to-image receptor distance; d
PID
: patient-to-image receptor distance.
Table C.7. Organ dose-conversion coefficients, i.e., D
T
/K
a,i
for different types of examination (Table C.6) for the CHILD
phantom.
Type of examination Skull AP Skull LAT Thorax AP Thorax PA Abdomen AP Pelvis AP
Body organs and tissues
Brain 0.285 0.351 0.003 0.003
a a
Eye lenses 1.30 0.563 0.006 0.002 0.002
a
Lungs 0.047 0.020 0.466 0.479 0.200 0.001
Ovaries
a a
0.003 0.001 0.569 0.505
Testes
a a a a
0.124 0.184
Thymus 0.173 0.060 0.616 0.238 0.036
a
Thyroid 0.417 0.585 0.799 0.163 0.012
a
Uterus
a a
0.001
a
0.627 0.636
Skeleton (hard bone) 0.694 0.691 0.351 0.502 0.449 0.309/0.303
b
Active BM 0.078 0.066 0.042 0.061 0.075 0.057
Total body 0.123 0.122 0.126 0.136 0.235 0.137/0.131
b
a
Less than 0.01.
b
The pelvic AP examination results for skeleton and total body in slightly higher doses for girls (first value) than for boys because of field
reduction due to shielding of testes.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
82
Table C.7 is K
a,i
D
T
c
T,K
a,i
K
a,i
: C:4
C.3.4 Illustrative calculations
Total active BM dose D
BM,tot
for BABY from an
abdomen AP examination performed with the same
physical exposure parameters as given in
Table C.4, but a K
a,i
(exposure at skin entrance)
value of only 43.0 mGy (5 mR)
D
BM;tot
5:83 mGy 43:0/76:2 3:3 mGy, C:5
where 5.83 mGy is the total active BM dose from
Table C.5, 43.0 mGy (5.0 mR) the actual K
a,i
(expos-
ure at skin entrance) value and 76.2 mGy (8.684 mR)
the K
a,i
(exposure at skin entrance) value from
Table C.4.
Total active BM dose D
BM
for CHILD from an
abdomen AP examination performed with the same
physical exposure parameters as listed in
Table C.6 and a K
a,i
value of 0.7 mGy is given by
D
BM
0:061 0:7 mGy 0:043 mGy, C:6
where 0.061 is the organ dose-conversion coefficient
for active BM from Table C.7.
C.4 REPORT ON ORGAN DOSES FROM CT
EXAMINATIONS (ADULTS) GSF-BERICHT
30/91 (ZANKL ET AL., 1991)
C.4.1 Main features of the calculations
The Monte Carlo calculations and the phantoms
used (ADAM and EVA) are the same as those for
radiographic examinations of adults (Section C.2).
The irradiation conditions for the calculation of
organ dose-conversion coefficients were derived
from a field study in Germany (Panzer et al., 1989).
Most of the CT scanners at that time were of the
following type.
Rotate only systems
Tube rotation 360

No beam shaping devices between focus and patient


Focus to axis of rotation distance 7076 cm
The radiation qualities considered are listed in
Table C.8.
C.4.2 Scope of the report
Table C.9 contains the organs and tissues for
which organ dose-conversion coefficients were
calculated. The doses to the active BM are given
both as dose to the active BM in the single bones
listed in Table C.9 and as the mean dose to the
total active BM distributed in the body. The dose
to the skeleton is given as the mean dose to the
whole skeleton, including both mineral bone and
active BM.
C.4.3 Application of the data
The organ dose-conversion coefficients c(organ, z)
presented in the Report express the organ dose
D(organ, z) resulting from a single axial 1 cm slice
with the axial co-ordinate z across the phantom per
unit of the CT kerma index C
K
. The z-axis has its
origin at the base of the trunk of each phantom. To
facilitate the positioning of a slice anatomical land-
marks with their axial co-ordinates z in the phan-
toms are provided.
D organ, z c organ, z C
K
: C:7
The normalization quantity C
K
should be based on
measurements at the respective unit using dosimet-
ers suitable for the measurement of C
K
and on
considerations of the influence of geometric effects
like overlapping of adjacent slices or spaces
between them. If measured dose values are not
available or if only rough estimates of organ doses
Table C.8. Radiation qualities in CT for which organ dose-
conversion coefficients were calculated.
Radiation quality
A B C
Tube voltage/kV 137 125 80
Total filtration/mm
a
2.2 Al 0.2 Cu
Mean photon energy/keV 66.0 64.4 50.5
HVL
1
/mm Al 8.06 7.98 5.38
a
The total filtration is the same for each of the spectra considered.
Table C.9. Organs and tissues for which CT organ dose-
conversion coefficients were calculated.
Adrenals Liver Skeleton
Bladder Lungs Spleen
Brain Ovaries Stomach
Breasts Pancreas Testes
Colon Total active bone marrow Thymus
Eye lenses Skin Thyroid
Kidneys Small Intestine Uterus
The bones for which the dose to the active bone marrow is listed
separately are the following.
Arm bones Leg bones Cervical spine
Clavicles Pelvis Thoracic spine
Cranium Ribs Lumbar spine
Facial skeleton Scapulae
APPENDIX C
83
are intended, values from literature can be used for
C
K
. Seven typical values for C
K
normalized to a P
It
of
100 mAs per slice or tube rotation are given in the
Report.
For each of the organs separate tables of organ
dose-conversion coefficients are given for ADAM
and EVA. A sample tabulation for the organ dose-
conversion coefficient for pancreas is given in
Table C.10. The first column indicates the z co-
ordinate of the slice. Each of the gender-specific
datasets contains also the percentage organ volume
included in the single slice, i.e., the fractions of
the organs being directly irradiated during
scanning by the respective 1-cm-thick slice. The
last column for each phantom presents an average
value of the coefficient of variation of the organ dose-
conversion coefficients estimated from the Monte
Carlo calculation.
The mean organ dose D
T
resulting for a complete
CT-examination is calculated by summing up the
values of the conversion coefficients c(organ, z) as
listed in the tables between the lower (z
l
) and the
upper (z
u
) boundary of the scanned region
D
T
C
K
c
T,C
K
C
K
X
z
u
z
l
c organ, z : C:8
In the case where a CT-examination consists of vari-
ous sections with different physical exposure para-
meters, the organ doses have to be calculated for
each section separately.
The method described above can also be applied
to examinations performed by spiral or multi-slice
CT, provided C
K
is correctly determined for these
techniques.
C.4.4 Illustrative calculation
Pancreas dose D
Pancreas
of the phantomEVAfor an
abdomen-CT consisting of 40 serial 8 mmslices (360

rotation, couch increment 8 mm) from thoracic ver-


tebra 12 (z co-ordinate in EVA: 41 cm) to the coccyx (z
co-ordinate: 9 cm). Radiation quality A from
Table C.8 and a P
It
of 200 mAs per slice
D
Pancreas
35:2 mGy 0:562 19:8 mGy, C:9
where the C
K
value of 35.2 mGy is either obtained
from either measurement, literature, or the table in
the Report, and the conversion coefficient of 0.562 is
the sum of the 32 c(organ, z) values listed in
Table C.10 from slice 9 to slice 41.
Table C.10. Sample table from Zankl et al. (1991) showing mean pancreas dose-conversion coefficients c(pancreas, z) for
single 1 cm axial CT slices and the radiation qualities A, B, C from Table C.8. The z coordinate of the slice is given in the
first column. The volume ratios of the organ located within each slice and the average coefficients of variation from the
Monte Carlo calculation are also given.
Slice
a
(cm)
Volume
ratio
%
Male phantom ADAM Coefficient of
variation
Volume
ratio
%
Female phantom EVA Coefficient of
variation
Conversion coefficients for
radiation qualities
Conversion coefficients for
radiation qualities
A B C A B C
4546 0.0105 0.0104 0.0073 0.031 0.0074 0.0069 0.0048 0.04
4445 0.0114 0.0115 0.0077 0.029 0.0080 0.0080 0.0051 0.038
4344 0.0144 0.0148 0.0103 0.026 0.0100 0.0098 0.0067 0.034
4243 0.0195 0.0194 0.0134 0.023 0.0121 0.0122 0.0079 0.031
4142 0.0220 0.0222 0.0157 0.022 0.0131 0.0137 0.0097 0.029
4041 1.0 0.0301 0.0290 0.0221 0.019 0.0187 0.0179 0.0137 0.024
3940 14.6 0.0515 0.0487 0.0375 0.016 0.0221 0.0226 0.0161 0.023
3839 26.6 0.0695 0.0639 0.0479 0.014 0.0280 0.0288 0.0207 0.02
3738 32.5 0.0817 0.0816 0.0624 0.013 8.6 0.0471 0.0462 0.0341 0.017
3637 9.7 0.0465 0.0461 0.0335 0.016 24.7 0.0716 0.0697 0.0523 0.014
3536 8.7 0.0382 0.0390 0.0274 0.018 33.5 0.0921 0.0903 0.0691 0.013
3435 6.2 0.0327 0.0323 0.0235 0.019 15.9 0.0623 0.0611 0.0452 0.015
3334 0.8 0.0240 0.0230 0.0169 0.021 9.5 0.0433 0.0440 0.0311 0.018
3233 0.0188 0.0183 0.0129 0.023 6.9 0.0382 0.0385 0.0281 0.018
3132 0.0151 0.0158 0.0106 0.025 0.9 0.0268 0.0271 0.0192 0.021
3031 0.0129 0.0124 0.0085 0.028 0.0209 0.0202 0.0149 0.023
2930 0.0104 0.0106 0.0068 0.031 0.0163 0.0166 0.0120 0.026
a
In the Report doses are given in this case for slices 71 down to 24.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
84
C.5 REPORT ON ORGAN DOSES FROM CT
EXAMINATIONS (PAEDIATRIC) GSF-BERICHT
30/93 (ZANKL ET AL., 1993)
C.5.1 Main features of calculations
The Monte Carlo program and the voxel phantoms
BABY and CHILD used to produce this catalogue of
organ dose-conversion coefficients were the same as
described in Section C.3.
The scanner types considered are rotate-only sys-
tems with a focus to axis of rotation distance
between 70 and 85 cm and angles of rotation of
360

or 180

, respectively. For 180

rotation, anterior
radiation incidence was modelled.
The use of beam-shaping devices between focus
and patient was rare at that time in Germany and
was, therefore, not modelled. Scans were, however,
often performed using an asymmetric beam, which
means that the beam width is reduced to 10 cm on
one side. Consequently, only a circle with a radius
of 10 cm around the axis of rotation is fully irradi-
ated, the area outside is not. Because of their
small body size, this makes no difference for babies.
For heavier children, such beams lead to markedly
lower doses in the parts of their body outside the
10 cm field. For these tissues, additional tables for
scans using asymmetric beams are given in the
catalogue.
The radiation qualities considered in the Report
are shown in Table C.11. They cover a wide energy
range and allow interpolation or, to some extent,
extrapolation for beam qualities deviating from
those given.
C.5.2 Scope of the report
The organs and tissues for which organ dose-
conversion coefficients were calculated are listed in
Table C.12. The doses to the active BM are
given both as dose to the active BM in the single
bones listed in Table C.12 and as the mean dose to
the total active BM distributed in the body. The dose
to the skeleton is given as the mean dose to the
entire skeleton, including both mineral bone
and BM.
C.5.3 Application of the data
The determination of organ doses follows the same
procedure as described in Section C.4.3 for adults
according to Eqs. (C.7) and (C.8). Also anatomical
landmarks within the phantoms and orientation val-
ues for C
K
normalized to a P
It
value of 100 mAs per
slice or tube rotation are presented.
For each of the organs considered, separate tables
of organ dose-conversion coefficients are given for
the BABY and for the CHILD phantom. A sample
tabulation of the data for the brain in the BABY
phantom resulting from 1 cm slices with a symmet-
rical beam is given in Table C.13. The first column
describes the single slices with respect to their z-co-
ordinate within the phantom. The next column
shows the percentage of the organ volume included
in the single slice, i.e., the ratio of the organ
being directly irradiated during scanning with a
1-cm-thick slice.
Two pairs of columns of conversion coefficients
follow, one pair for each of the two radiation qualit-
ies considered. Within each of these pairs, the first
column presents conversion coefficients for full 360

rotation of the fan beam, the second column those for


180

rotation around the anterior surface of the


body. The last column in each table presents an
average value of coefficient of variation of the
organ dose-conversion coefficients estimated from
the Monte Carlo calculation.
C.5.4 Illustrative calculation
Organ dose to brain of the BABY phantom from a
CT skull examination consisting of 12 contiguous
1-cm-thick slices from the top of the skull performed
Table C.11. Radiation qualities use for the calculation of
organ dose-conversion coefficients for paediatric CT-
examinations.
Radiation quality
Tube voltage/kV 80 125
Total filtration/mm
a
2.2 Al 0.2 Cu
Mean photon energy/keV 50.5 64.4
HVL
1
Al 5.38 7.98
a
The total filtration is the same for each of the spectra considered.
Table C.12. Organs for which organ dose-conversion
coefficient was calculated for paediatric CT-examina-
tions.
Adrenals Liver Spleen
Bladder Lungs Stomach
Brain Ovaries Testes
Breasts Pancreas Thymus
Colon Active BM Thyroid
Eye lenses Skeleton Tissue
Oesophagus Skin Uterus
Kidneys Small intestine
The bones for which the dose to the active BM is listed separately
are:
Arm bones Mandible Cervical spine
Clavicles Pelvis Thoracic spine
Cranium Ribs Lumbar spine and sacrum
Leg bones Scapulae Sternum
APPENDIX C
85
with 125 kV tube voltage, 2.2 mm Al 0.2 mm Cu
filtration, a P
It
of 100 mAs per slice, 360

tube rota-
tion and symmetrical beam:
D
Brain
14:4 mGy 0:902 13 mGy, C:10
where C
K
is 14.4 mGy from measurement or
from literature values given in the Report and
0.902 the sum of the 12 c(organ, z) values listed in
Table C.13 from slice 0 cm to 1 cm until slice 11 cm to
12 cm.
Table C.13. Sample table from Zankl et al. (1993) showing mean brain dose coefficients (D
T
/C
K
for a series of transverse CT
slices of the BABY phantom and average coefficients of variation per single 1-cm-thick slice. The volume ratio of the organ
located within each single slice is also shown.
Slice (cm) Volume ratio (%) Conversion coefficients Coefficient of variation
80 kV 125 kV
360

180

360

180

01 2.4 0.0276 0.0287 0.0314 0.0322 0.0034


12 8.1 0.0685 0.0711 0.0795 0.0813 0.0023
23 13.7 0.102 0.1028 0.1178 0.120 0.0019
34 16.9 0.119 0.120 0.140 0.141 0.0017
45 18.1 0.124 0.125 0.146 0.147 0.0017
56 17.1 0.116 0.115 0.1361 0.1361 0.0017
67 12.4 0.0886 0.0838 0.105 0.101 0.0020
78 7.5 0.0593 0.0550 0.0709 0.0667 0.0024
89 3.3 0.0319 0.0289 0.0398 0.0369 0.0031
910 0.4 0.0140 0.0128 0.0181 0.0170 0.0045
1011 0.0084 0.0082 0.0108 0.0108 0.0057
1112 0.0054 0.0052 0.0071 0.0071 0.0071
1213 0.0038 0.0038 0.0051 0.0052 0.0084
1314 0.0027 0.0027 0.0037 0.0037 0.0099
1415 0.0021 0.0022 0.0030 0.0030 0.0109
1516 0.0016 0.0017 0.0024 0.0024 0.0124
1617 0.0012 0.0013 0.0018 0.0019 0.0141
1718 0.0009 0.0010 0.0014 0.0015 0.0162
1819 0.0007 0.0007 0.0011 0.0011 0.0185
1920 0.0005 0.0005 0.0008 0.0008 0.0211
2021 0.0003 0.0004 0.0006 0.0006 0.0254
2122 0.0003 0.0005 0.0005 0.0288
2223 0.0002 0.0002 0.0004 0.0004 0.0308
2324 0.0002 0.0002 0.0003 0.0003 0.0343
2425 0.0002 0.0002 0.0003 0.0003 0.0383
2526 0.0001 0.0001 0.0002 0.0002 0.0433
2627 0.0001 0.0001 0.0002 0.0002 0.0490
2728 0.0001 0.0001 0.0001 0.0001 0.0556
2829 0.0001 0.0001 0.0001 0.0001 0.0634
2930 * * 0.0001 0.0001 0.0721
3031 * * 0.0001 0.0001 0.0835
An asterisk as table entry means that the organ dose-conversion coefficient is <0.00005.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
86
Journal of the ICRU Vol 5 No 2 (2005) Report 74 doi:10.1093/jicru/ndi032
Oxford University Press
APPENDIX D: REPORTS PRODUCED BY THE HEALTH
PROTECTION AGENCY (HPA) (FORMERLY NATIONAL
RADIOLOGICAL PROTECTION BOARD, NRPB)
D.1 GENERAL DESCRIPTION
NRPB has produced three sets of reports that
permit estimation of the mean absorbed doses to
25 different tissues and organs in reference adult
and paediatric patients from common radiographic,
fluoroscopic, and CT examinations.
The three sets of reports deal with the following.
Each set consists of a pair of NRPB reports, one
providing printed tabulations of conversion coeffi-
cients for a sample of the available data and the
other (the software report, denoted NRPB-SR. . .)
includes a computer disk containing a complete list-
ing of all the available dose-conversion coefficients.
The listing is arranged in a large series of data files,
which can be incorporated into suitable computer
programs for processing the conversion coefficients
in a manner matched to the requirements of the
user. Such user-friendly software packages, which
calculate organ doses for any combination of x-ray
examination parameters specified by the user, are
available from the National Radiation Laboratory,
Christchurch, New Zealand, for each set of NRPB
reports.
The dose-conversion coefficients in the software
reports relate organ dose D
T
/mGy to entrance
surface air kerma, K
a,e
/mGy and to air kermaarea
product, P
KA
/(Gy cm
2
) for radiographic and
fluoroscopic examinations and to CT kerma index
expressed in ICRU muscle tissue, C
K,m
/mGy for CT
examinations.
The same basic Monte Carlo code developed at
NRPB is used for all three sets of calculations and
both adult and paediatric patients are simulated by
hermaphrodite mathematical phantoms. The soft-
ware reports include organ dose-conversion coeffi-
cients for a large number of diagnostic radiation
qualities or types of CT scanner.
D.2 REPORT NRPB-SR262 (HART ET AL.,
1994b): COMMON ADULT X-RAY
EXAMINATIONS
D.2.1 Main features of calculations
Details of the Monte Carlo code and the mathe-
matical phantom developed at NRPB are described
by Jones and Wall (1985). Conversion coefficients
relating D
T
to K
a,e
and to P
KA
were calculated
by simulating typical x-ray examination exposure
conditions on a hermaphrodite mathematical phan-
tom representing a typical adult patient. Four mil-
lion photon histories were followed for each x-ray
examination field simulated. The phantom was
based on Cristys adult hermaphrodite phantom
(Cristy, 1980) with Kramers modified neck region
(Kramer et al., 1982), a breast tissue composition of
50 %fat and 50 %water, and with an esophagus. The
arms of the phantom, which lie alongside the trunk,
were removed when lateral projections through the
trunk were simulated.
D.2.2 Scope of the report
Report NRPB-SR262 (Hart et al., 1994b) is a soft-
ware report, which includes a computer disk con-
taining files of conversion coefficients relating D
T
to K
a,e
and to P
KA
for the 68 radiographic projection
(e.g., Head) and view (e.g., AP) combinations listed
below.
1. Common Adult X-ray
Examinations
(NRPB-R262,
NRPB-SR262,
Hart et al., 1994a,b)
2. Common Paediatric
X-ray Examinations
(NRPB-R279,
NRPB-SR279,
Hart et al., 1996a,b)
3. Adult Computed-
tomography
Examinations
(NRPB-R250,
NRPB- SR250
Jones and Shrimpton,
1991, 1993)
International Commission on Radiation Units and Measurements 2005
The focus-to-surface distance d
FSD
, the position of
the central axis of the x-ray beam in the phantom,
and the x-ray field size at the image receptor and at
the mid-plane of the phantom, for all the above
radiographic projections are described in the report.
The mid-plane field size is also shown diagrammat-
ically on a drawing of the phantom so that the posi-
tion of the x-ray field in relation to all the modelled
organs can be visualized.
For each radiographic projection and view, conver-
sion coefficients are tabulated for 40 x-ray spectra,
specified by the peak tube voltage and the total beam
filtration. Tube voltages range from 50 to 120 kV, in
10 kVsteps; and five total beamfiltrations of 2, 2.5, 3,
4, and 5 mm Al are included for each tube voltage.
The source of the x-ray spectra data for these calcu-
lations is an updated (1991) version of the computer
programof Iles (1987). This programproduces results
very similar to that published by the IPEM (1997).
In NRPB-SR262 (Hart et al., 1994b) conversion
coefficients are given for 26 organs for the 68 radio-
graphic projection and view combinations and 40
x-ray spectra described above. These 26 organs are
the following.
D.2.3 Application of the report
The information required to estimate organ doses
from 68 adult radiographic projections and views
using the conversion coefficients listed in NRPB-
SR262 is K
a,e
or P
KA
for each radiograph and the
corresponding tube voltage and total beamfiltration.
Organ doses can be estimated for any combination
of the 68 radiographic projections and views, using a
software package called XDOSE developed at the
National Radiation Laboratory, Christchurch, New
Zealand. This software processes the data in NRPB-
SR262 according to the requirements of the user. An
example of part of the results produced by this soft-
ware for a PA chest radiograph and an AP abdom-
inal radiograph with specified exposure factors is
shown in Table D.1.
D.2.4 Illustrative calculation
Organ doses D
T
for chest PA and abdomen AP
examinations were calculated using NRPB-SR262
Monte Carlo data (Hart et al., 1994b). The chest PA
examination is performed with a tube voltage of 120
kV, a filtration of 3.00 mm Al, and at the reference
d
FSD
. The entrance air kerma K
a,e
for the chest exam-
ination is 0.16 mGy. The abdomen AP examination is
performed with a tube voltage of 85 kV, a filtration of
3.00 mm Al, and at the reference d
FSD
. The entrance
air kerma K
a,e
for the abdomen examination is 5.60
mGy. The results of the organ dose calculation are
given in Table D.1.
D.3 REPORT NRPB-SR279 (HART ET AL.,
1996b): COMMON PAEDIATRIC X-RAY
EXAMINATIONS
D.3.1 Main features of calculations
Details of the Monte Carlo code developed at
NRPB are described by Jones and Wall (1985). Con-
version coefficients relating D
T
to K
a,e
and to P
KA
were calculated by simulating typical x-ray exam-
ination exposure conditions on a series of five herm-
aphrodite mathematical phantoms representing
newborn, 1, 5, 10, and 15-year-old patients. Four
million photon histories were followed for each x-
ray examination field that was simulated. The
Adrenals Lungs Spleen
Brain Oesophagus Stomach
Breasts Ovaries Testes
Eye Lens Pancreas Thymus
Gall bladder Active BM Thyroid
Heart
Kidneys
Residual tissues
(muscle)
Upper large
intestine
Liver
Lower large
Skeleton (bone
surfaces)
Urinary
bladder
intestine Skin
Small intestine
Uterus
Head AP, PA Heart LAT L and R Duodenum AP, PA
Head LAT L and R Heart LAO, RAO Duodenum RAO, LPO
Cervical spine AP Upper stomach AP, PA LSJ LAT L and R
Cervical spine LAT L and R Stomach RAO, LPO, LAO Abdomen AP, PA
Throat LAT L and R Stomach AP, PA, Small intestine AP, PA
Shoulder AP L and R Stomach LAT (L) Pelvis/colon AP, PA
Oesophagus RAO, LAO, LPO Kidneys AP, PA Colon RAO, LPO, LAO
Thoracic spine AP Lumbar spine LPO, RPO Urinary bladder AP
Thoracic spine LAT L and R Lumbar spine AP Hip AP L and R
Chest AP, PA Lumbar spine LAT L and R Rectum AP, PA
Chest LAT L and R Left flexure LAO Rectum LAT L and R
Heart AP, PA Right flexure RAO Rectum LPO, RAO
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
88
phantoms were based on Cristys hermaphrodite
phantoms (Cristy, 1980) and Kramers modified
neck region (Kramer et al., 1982) was applied to all
of them. An esophagus was added to each phantom.
A suitable volume of breast tissue was included in
each phantom and in the absence of specific informa-
tion on the fat content of breast tissue in young
children, 50 % fat and 50 % water was used as the
breast tissue composition for all of the phantoms.
The arms of the phantom, which lie alongside the
trunk, were removed when lateral projections
through the trunk were simulated.
The weights and major dimensions of the five
paediatric phantoms are shown below.
D.3.2 Scope of the report
Report NRPB-SR279 (Hart et al., 1996b) contains
files of conversion coefficients relating D
T
to K
a,e
and
to P
KA
for the 18 radiographic projections and views
listed below.
The d
FSD
, the position of the central axis of the
x-ray beam in each phantom, and the x-ray field size
at the image receptor, for all the above radiographic
projections are described in the report. The field size
at the mid-plane of the phantom is also shown dia-
grammatically on a drawing of each phantom so that
the position of the x-ray field in relation to all the
modelled organs can be visualized.
For each radiographic projection and view, con-
version coefficients are tabulated for 72 x-ray
spectra, specified by the peak tube voltage and
the total beam filtration. Tube voltage ranges from
50 to 120 kV, in 10 kV steps. For each tube
voltage nine total beam filtrations are included, the
same five as for the previous adult examinations
(i.e., 2, 2.5, 3, 4, and 5 mm Al) and a further four
involving additional copper filtration that has been
recommendedfor paediatric radiography(EC1996d),
i.e., 2 mm Al 0.1 mm Cu; 2 mm Al 0.2 mm Cu;
3 mm Al 0.1 mm Cu and 3 mm Al 0.2 mm Cu.
The source of the x-ray spectra data for these calcu-
lations is an updated (1991) version of the computer
program of Iles (1987). This program produces very
similar results to that published by the IPEM (1997).
In NRPB-SR279 (Hart et al., 1996b) conversion
coefficients for 26 organs are given for the 18 radio-
graphic projections and views and the 72 x-ray
spectra described above. These organs are the
following.
Age/
years
Weight/
kg
Height/
cm
Thickness
of head/cm
Thickness
of trunk/
cm
AP LAT AP LAT
0 3.47 51.5 11.6 9.0 9.8 12.7
1 9.26 75.0 15.7 12.3 13.0 17.6
5 19.0 109.0 18.1 14.3 15.0 22.9
10 31.9 138.6 18.8 14.9 16.8 27.8
15 54.4 164.0 19.5 15.5 19.6 34.5
Head AP,
PA, LAT
Lumbar spine
AP, PA, LAT
Urinary
bladder
Cervical spine
AP, PA, LAT
Abdomen AP, PA
Pelvis AP, PA
AP, PA
Chest AP,
PA, LAT
Adrenals Lungs Spleen
Brain Oesophagus Stomach
Breasts Ovaries Testes
Eye lens Pancreas Thymus
Gall bladder Active BM Thyroid
Heart
Kidneys
Residual
tissues (muscle)
Upper
large intestine
Liver
Lower large
Skeleton
(bone surfaces)
Urinary bladder
Uterus
intestine Skin
Small intestine
Table D.1. Organ doses D
T
/mGy calculated for chest PA
and abdomen AP examinations using NRPB-SR262 Monte
Carlo data and the software package XDOSE for the
examination conditions given in Section D.2.4.
Organ D
T
/mGy
Chest PA Abdomen AP
Adrenals 0.0970 0.261
Brain 0.0006 0.00
Breasts 0.0197 0.0199
Eye lens 0.0003 0.00
Gall bladder 0.0187 1.95
Stomach 0.0238 1.95
Small intestine 0.0025 1.64
Upper large intestine 0.0032 1.98
Lower large intestine 0.0006 1.32
Heart 0.0362 0.0675
Kidneys 0.0608 0.366
Liver 0.0419 1.08
Lungs 0.0846 0.0474
Ovaries 0.0005 1.28
Pancreas 0.0453 0.872
Skin 0.0163 0.387
Spleen 0.0803 0.507
Testes 0.00 0.243
Thymus 0.0184 0.0094
Thyroid 0.0108 0.0003
Urinary bladder 0.0001 2.61
Uterus 0.0005 1.67
Oesophagus 0.0490 0.0643
Muscle 0.0181 0.531
Total bone 0.0485 0.289
Active BM 0.0274 0.217
APPENDIX D
89
D.3.3 Application of the report
The information required to estimate organ
doses from 18 paediatric radiographic examinations
using the conversion coefficients listed in NRPB-
SR279 is the K
a,e
or the P
KA
for each radiograph
and the corresponding tube voltage and total beam
filtration.
Organ doses to any of the phantoms representing
a newborn, 1, 5, 10, or 15-year-old patient can be
estimated for any combination of the 18 radiographic
projections, using a software package called
CHILDOSE developed at the National Radiation
Laboratory, Christchurch, New Zealand. This soft-
ware processes the data in NRPB-SR279 according
to the requirements of the user. An example of
part of the results produced by this software for an
examination comprising a PA chest radiograph and
an AP abdominal radiograph on a 1-year-old child
with the specified exposure factors is shown in
Table D.2.
D.3.4 Illustrative calculation
Organdoses D
T
for chest PAand abdomenAPexam-
inations were calculated using NRPB-SR279 Monte
Carlo data (Hart et al., 1996b) for child of 1 year.
The chest PA examination is performed with a tube
voltage of 90 kV, a filtration of 2.00 mmAl and of 0.10
mm Cu, and at the reference d
FSD
. The entrance air
kerma K
a,e
for the chest examination is 0.05 mGy.
The abdomen AP examination is performed with a
tube voltage of 90 kV, a filtration of 2.00 mm Al and
0.10 mmCu, and at the reference d
FSD
. The entrance
air kerma K
a,e
for the abdomen examination is 0.40
mGy. The results of the organ dose calculation are
given in Table D.2.
D.4 REPORTS NRPB-R250, NRPB-SR250
(JONES AND SHRIMPTON, 1991, 1993): ADULT
CT EXAMINATIONS
D.4.1 Main features of calculations
The same Monte Carlo code and adult hermaph-
rodite mathematical phantom were used as in the
calculations for conventional x-ray examinations
(see Appendix D.2.1). Only the esophagus was miss-
ing from the phantom described in Appendix D.2.1.
The arms of the phantom remained alongside the
trunk for all simulated CT examinations, as is the
usual practice with patients.
To simulate the distinct exposure conditions pre-
vailing during CT examinations, the x-ray source
was rotated around the long axis of the phantom.
Organ doses were determined for the individual irra-
diation of 208 contiguous 5-mm-thick transverse
slabs of the phantom ranging from the top of the
head to the top of the legs. A line x-ray source of
the same length of each irradiated slab (5 mm) was
used, with the photons emitted normally from the
source into a fan-shaped beam spanning the width of
the phantom. An abrupt cut-off of the x-ray beam at
the edges and faces of the fan (i.e., perfect collima-
tion) is a feature of this model. Conversion coeffi-
cients were calculated relating the organ doses to
the kerma on the axis of rotation of the scanner
free-in-air, i.e., with no phantom or patient present.
The imperfect collimation of the x-ray beam in the
axial direction, which occurs in practice, can lead to
a build-up of dose in adjacent scans. However, this is
taken account of by measuring the CT kerma index
free-in-air, C
K
, under the exposure conditions, i.e.,
slice thickness, tube voltage, and tube-current
exposure-time product, P
It
. The value of C
K
is then multiplied by the conversion coefficients
to estimate the organ doses for the degree of
beam collimation achieved by the CT scanner of
Table D.2. Organ doses D
T
/mGy for child of 1 year
calculated for chest PA and abdomen AP examinations
using NRPB-SR279 Monte Carlo data and CHILDOSE
software for the examination conditions given in
Section D.3.4.
Organ D
T
/mGy for child of 1 year
Chest PA Abdomen AP
Adrenals 0.0215 0.0796
Brain 0.0002 0.0001
Breasts 0.0105 0.0102
Eye lens 0.0002 0.0002
Gall bladder 0.0026 0.252
Stomach 0.0049 0.271
Small intestine 0.0008 0.211
Upper large intestine 0.0009 0.242
Lower large intestine 0.0002 0.152
Heart 0.0144 0.0440
Kidneys 0.0029 0.0796
Liver 0.0081 0.226
Lungs 0.0311 0.0390
Ovaries 0.0004 0.184
Pancreas 0.0079 0.167
Skin 0.0063 0.0468
Spleen 0.0107 0.118
Testes 0.00 0.0335
Thymus 0.0080 0.0065
Thyroid 0.0060 0.0024
Urinary bladder 0.0001 0.264
Uterus 0.0003 0.216
Oesophagus 0.0177 0.0508
Muscle 0.0068 0.0748
Total bone 0.0259 0.0860
Active BM 0.0060 0.0274
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
90
interest. In these early (pre-1990) calculations
kerma calculated on the axis of rotation of the scan-
ner was expressed in terms of ICRU muscle tissue,
K
m
, not as air kerma K
a
. Consequently the C
K
should
be converted into CT kerma index expressed in
terms of absorbed dose to ICRU muscle, C
K,m
, when
using these conversion coefficients to estimate organ
doses.
Conversion coefficients relating organ doses to
C
K,m
measured free-in-air on the scanner axis for
any complete CT examination of the head or trunk
can be obtained by suitable summation of the con-
version coefficients for the irradiation of the indi-
vidual 5-mm-thick slabs.
D.4.2 Scope of the reports
Calculations have been performed for 23 sets of
exposure conditions (i.e., tube voltage, focus-to-axis
distance, and total filtration), selected to cover 27
models of CT scanner from five different manufac-
turers, which were in common use in the UK in the
late 1980s (Table D.3). Shaped (bow-tie) filters
have been modelled where necessary. The exposure
conditions assumed for the 23 sets of calculations
and the appropriate CT scanner models are listed
below.
Report NRPB-R250 (Jones and Shrimpton, 1991)
contains tables of conversion coefficients relating
doses to 27 organs or regions of the phantom to the
C
K,m
for each of the 208 5 mm slabs and for four of
the datasets listed below. Report NRPB-SR250 is a
software report that contains a computer disk on
which the conversion coefficients for all the datasets
listed below are recorded.
The 27 organs and regions of the phantom for
which conversion coefficients have been calculated
are the following.
Table D.3. Exposure conditions for calculation of normalized organ doses for CT using Monte Carlo techniques according
to Jones and Shrimpton (1991, 1993).
Monte Carlo
data set no.
CT scanner models for which
data set is appropriate
Tube
voltage/kV
Focus to axis
distance/mm
Flat filter/ Shaped filter/
Al mm Cu mm Option Type
Siemens
1 Somatom 2,DR1/2/3 125 760 2.2 0.25
2 Somatom DRG,DRG1 125 700 2.2 0.4
3 Somatom DRH,CR,CR512 125 700 2.2 0.2
Picker
4 1200SX 130 640 0.7 HEAD LEXAN
a
5 1200SX 130 640 0.7 BODY LEXAN
a
GE
6 CT 8800,9000I/II/HP 120 780 2.7 HEAD PMMA
7 CT 8800,9000I/II/HP 120 780 2.7 BODY PMMA
8 CT 9800,9800Quick 120 630 2.7 PTFE
b
9
c
CT 9800,9800Quick 140 630 2.7 PTFE
b
10 CT MAX 120 525 2.6 PTFE
b
11 CT PACE 120 525 2.7 PTFE
b
CGR
12 CE 10000,12000 130 750 1.0 0.3 PEEK
d
Philips
e
13 Tomoscan 305N,310,350 (GE No. 2, no Cu) 120 487 3.5 Al
14 Tomoscan 305N,310,350 (GE No. 2, with Cu) 120 487 3.5 0.25 Al
15 Tomoscan 310,350 (GE No.3, no Cu) 120 608 3.5 Al
16 Tomoscan 310,350 (GE No. 3, with Cu) 120 608 3.5 0.25 Al
17 Tomoscan TX 120 606 1.4 0.1 Al
18 Tomoscan CX,CX/S 120 606 1.4 0.1 Al
19 Tomoscan LX 120 606 1.4 0.1 Al
20
c
Tomoscan TX 100 606 1.4 0.1 Al
21
c
Tomoscan TX 130 606 1.4 0.1 Al
22
c
Tomoscan LX 100 606 1.4 0.1 Al
23
c
Tomoscan LX 130 606 1.4 0.1 Al
a
LEXAN is polycarbonate.
b
PTFE is polytetrafluoroethylene.
c
Datasets 9, 20, 21, 22, and 23 relate to alternative applied potential settings for selected scanners.
d
PEEK is polyetheretherketone.
e
Philips Tomoscan 300 series scanners have options for the variation of focus-to-axis distance (geometric enlargement setting) and for the
inclusion of additional copper filtration.
APPENDIX D
91
D.4.3 Application of the reports
The information required to estimate organ doses
for CT examinations using the conversion coeffi-
cients tabulated in NRPB-R250 and listed in
NRPB-SR250 is C
K,m
for the particular scanning
parameters in clinical use (tube voltage, P
It
, slice
thickness, focus-to-axis distance, and filtration).
Linearity of C
K,m
with P
It
is assumed.
The conversion coefficients need to be summed for
all of the 5-mm-thick slabs that lie within the
scanned volume for a particular CT examination.
The value of D
T
to a given organ from the CT slices
imaged within a scanned volume is given by
D
T
C
K;m
p
X
z
u
z
l
c
T,C
K;m
, D:1
where p is the packing factor and the conversion
coefficients for the required organ are summed
between the lower (z
l
) and upper (z
u
) boundaries of
the scanned volume. The packing factor takes
account of the relationship between slice thickness
and table increment during a sequence of serial
slices (or the pitch during a helical scan).
For serial scans
p
wn
L
, D:2
where wis the slice thickness, n the number of slices,
L the total length of the scan, and p 1 for contigu-
ous slices, p > 1 for overlapping slices, and p < 1 for
gaps between slices.
For helical scans, p the pitch.
Organ doses can be estimated for any CT exam-
ination of the head or trunk, using a software pack-
age called CTDOSE developed at the National
Radiation Laboratory, Christchurch, New Zealand.
This software processes the data in NRPB-SR250
according to the requirements of the user. An
example of part of the results produced by this soft-
ware for a CT examination of the chest on a Philips
Tomoscan LX scanner with just one series of scans
(one region) and the specified scanning parameters
is shown in Section D4.4. CTDOSE allows up to four
series of scans (regions) involving the use of differ-
ent slice thickness, P
It
, and table increment com-
binations in any one CT examination. With no
esophagus modelled in the phantom, the dose to
the esophagus is taken to be the same as the dose
to the thymus.
D.4.4 Illustrative calculation
Organ doses D
T
for routine chest CT examina-
tions, region 1, according to NRPB-R250 combined
with CTDOSE. The scan was performed with a tube
voltage of 125 kV, P
It
of 350 mAs, C
K,m
/P
It
of
0.220 mGy (mAs)
1
, slice thickness of 10 mm,
table increment of 10 mm, number of slices 20,
start position of 430 mm, and end position of 650
mm. The resulting organ doses are given in
Table D.4.
Adrenals Lungs Testes
Brain Ovaries Thymus
Breasts Pancreas Thyroid
Eye lens Active BM Upper large
Gall bladder
Heart
Skeleton
(bone surfaces)
intestine
Urinary bladder
Kidneys Skin Uterus
Liver Small intestine Head region
Lower large Spleen Trunk region
intestine Stomach Leg region
Table D.4. Organ doses D
T
/mGy for routine chest CT
examination, region 1, using NRPB-SR250 Monte Carlo
data and CTDOSE software for the examination condi-
tions given in Section D.4.4.
Organ D
T
/mGy
Adrenals 2.97
Brain 0.08
Breasts 17.50
Eye lens 0.07
Gall bladder 0.76
Stomach 1.55
Small intestine 0.10
ULI 0.14
LLI 0.02
Oesophagus 31.90
Heart 23.10
Kidneys 0.61
Liver 2.37
Lungs 20.60
Ovaries 0.02
Pancreas 2.30
Skin 3.24
Spleen 1.75
Testes 0.0004
Muscle 3.73
Thymus 31.90
Thyroid 1.86
Urinary bladder 0.005
Uterus 0.029
Head region 0.62
Trunk region 7.86
Leg region 0.0002
Bone 9.76
ABM 5.05
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
92
Journal of the ICRU Vol 5 No 2 (2005) Report 74 doi:10.1093/jicru/ndi033
Oxford University Press
APPENDIX E: REVIEW OF MONTE CARLO CALCULATIONS
FOR ASSESSMENT OF MEAN GLANDULAR DOSE IN
MAMMOGRAPHY
E.1 GENERAL
Among the quantities used for dose specification
in mammography the average absorbed dose in
glandular tissue D
G
is most appropriate for risk
assessments (NCRP, 1986). Generally, D
G
is derived
from incident air kerma K
a.i
combined with conver-
sion coefficients obtained from radiation transport
calculations in mathematical models of the breast.
For simplicity the notation c
G
Eq. (E.1) is used
instead of c
G,K
a,i
Eq. (3.3.8)
D
G
c
G
K
a;i
: E:1
Conversion coefficients c
G
have been published by
various authors (Stanton et al., 1984; Rosenstein
et al., 1985; Dance, 1990; Wu et al., 1991, 1994;
Alm Carlsson and Dance, 1992; Jansen et al., 1994;
Zoetelief and Jansen, 1995; Klein et al., 1997;
Dance et al., 2000). Values of c
G
are usually presen-
ted as a function of the HVL
1
of the incident radi-
ation. Commonly, c
G
values are calculated using
simple breast phantoms. These breast models con-
tain a superficial layer of 0.4 or 0.5 cm thickness
representing the skin and underlying adipose tissue
and a central region consisting of a mixture of
adipose and glandular tissue (Figure E.1). Most of
the publications deal with firm compression of the
breast only, because it has been clearly demon-
strated that this is mandatory in view of dose reduc-
tion and improvement of image quality (NCRP,
1986).
Concerning the elemental tissue composition it
should be noted that most authors employ data pub-
lished by Hammerstein et al. (1979). The influence of
the variation in the fraction of glandular tissue in
the central region of the simulated female breast is
included in several publications (Rosenstein et al.,
1985; Wu et al., 1991a, 1994; Jansen et al., 1994;
Klein et al., 1997; Dance et al., 2000). Alm Carlsson
and Dance (1992) and Zoetelief and Jansen (1995)
studied the influence of variations in elemental com-
position of glandular and adipose tissues for the
range of values stated by Hammerstein et al. (1979)
and between Hammerstein et al. (1979) and ICRU
(1989) elemental compositions, respectively.
The calculations by different authors differ in
radiation transport codes, photon interaction data,
photon spectra, composition and thickness of super-
ficial layer (representing skin and subcutaneous
adipose tissue), presence of a compression plate,
and focus-to-film distance d
FFD
. They are generally
not performed with the tissue compositions recom-
mended by the ICRU (1989a). A study on the influ-
ence of these differences on c
G
(Table E.1) was made
by Zoetelief and Jansen (1995). It can be concluded
from the table that there are various sources of
16 cm
Plan
0.4 cm or 0.5 cm
variable
thickness
Adipose tissue
50 : 50 Adipose and glandular tissue
Elevation
Figure E.1. Mathematical model of the breast used in Monte
Carlo calculations of average glandular dose. Shown are the
superficial layer (either 0.4 cm of glandular tissue or 0.5 cm of
adipose tissue) and the central region consisting of 50 % adipose
and 50 % glandular tissue by mass (Reproduced from Dance,
1990, with permission from IOP, UK). The 50:50 breast model
was first suggested by Hammerstein et al. (1979).
International Commission on Radiation Units and Measurements 2005
difference with values ranging from 3 to 19 %.
Consequently, differences in c
G
-values of the order
of 20 to 30 % could have occurred between results
published by different authors. However, in practice
maximum differences of 15 % are observed.
E.2 SCOPE
Three approaches for dose assessment in mammo-
graphy are presented in this appendix. The first
method (Appendix E.3) concerns determination of
D
G
for a 4.5 (ACR, 1992) or 5-cm-thick standard
breast (EC, 1996b). According to the American Col-
lege of Radiology (ACR) protocol, K
a,i
is measured
using a mammographic phantom equivalent to
4.5 cm compressed breast tissue. The EC recom-
mends K
a,i
measurements for 5-cm-thick breasts or
alternatively for a 4.5-cm-thick PMMA phantom
(used in practice to mimic a standard breast). D
G
then results from measured K
a,i
and application of
the appropriate conversion coefficient c
G
, selected
on the basis of HVL
1.
The second method (Appendix E.4) also takes
into account the variation in compressed breast
thickness. The method does not allow for corrections
due to variations in the fraction of glandular tissue
present in the breast. This method will be advant-
ageous when the distribution of compressed breast
thickness cannot be reasonably represented by
standard breast thicknesses of 4.5 or 5 cm.
The third method (Appendix E.5) enables assess-
ment of D
G
for variations in compressed breast
thickness and fraction of glandular tissue. This
method requires an assessment of the fraction of
glandular tissue.
E.3 CONVERSION COEFFICIENTS C
G
FOR A
4.5 OR 5-CM-THICK STANDARD BREAST
Table E.2 shows a summary of conversion coeffi-
cients c
G
calculated by various authors for a stand-
ard sized breast, i.e., having a total thickness of 5 cm
and a central region consisting of 50 % adipose and
50 % glandular tissue by mass, examined under firm
compression. At an HVL
1
of 0.30 mm Al employing
a Mo/Mo anode-filter combination the largest differ-
ence is 12 %; at values of HVL
1
of 0.35 and 0.4 mm
Al for the same anode/filter combination these
Table E.2. Conversion coefficient c
G
that relates K
a,i
to D
G
for a 5-cm-thick breast having a central region of 50 % adipose
and 50 % glandular tissue by mass (tissue compositions according to Hammerstein et al., 1979).
HVL
1
Rosenstein et al.
(1985)
a
Dance
(1990)
b
Wu et al.
(1991)
Jansen et al. (1994)
c
Wu et al. (1994)
d
Dance et al.
(2000)
e
mm Al (Mo/Mo)
0.30 0.171 (0.160) 0.164 0.146 0.1510.153 0.164
0.35 0.187 0.172 0.168 (Mo/Mo) 0.1740.180 0.1880.194
0.40 0.217 (0.200) 0.209 0.195 0.1910.198 (W/MoMo/Mo) 0.1960.209 0.2160.224
0.45 0.232 0.2210.229 (Mo/RhRh/Rh) 0.2200239 0.2410.253
0.50 0.258 0.257 (Rh/Rh) 0.2420.260 0.270
0.55 0.287 0.273 (W/Rh) 0.296
0.60 0.314 0.310 0.294 (W/Rh) 0.321
0.65 0.332
a
Conversion coefficients for tungsten targets; conversion coefficients given in brackets refer to Mo and Mo/W targets
b
HVL
1
range: 0.250.45 mm Al: Mo/Mo; 0.450.70 mm Al: W/Mo; 0.500.80 mm Al: W/Rh; 0.550.90 mm Al: W/Pd; 0.502.00 mm Al: W/Al.
c
Anode/filter materials include Mo/Mo, W/Mo, Mo/Rh, Rh/Rh and W/Rh.
d
Lower values refer to Mo/Rh and higher values to Rh/Rh. Values can differ by 5 % depending on differences in tube voltage.
e
Anode filter combinations also include Mo/Mo, Mo/Rh, Rh/Al, Rh/Rh and W/Rh (Dance, 1990).
Table E.1. Influence of difference in approach by various authors in the calculation of c
G
(Zoetelief and Jansen, 1995).
Parameter Range of variation
Spectra at same HVL
1
for same tube voltage and anode-filter combination 7 %
Photon interaction data: MCPLIB (1988) versus XCOM (Berger and Hubbell, 1987) 10 %
Presence or absence of compression plate 34 % (3 mm PMMA)
Composition and thickness of superficial layer Hammersteins tissue composition 1119 % (breast thickness 8-2 cm)
ICRU tissue composition 310 % (breast thickness 8-2 cm)
Tissue composition Hammerstein versus ICRU 1114 % (at breast thickness 8-2 cm)
Fraction of glandular tissue in central region (HVL
1
: 0.4 mm Al, 6-cm-thick breast, 25 %
glandular tissue versus 50 % glandular tissue)
1013 %
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
94
differences are reduced to 9 and 7 %, respectively.
For larger values of HVL
1
the differences remain
approximately constant at 7 %. These differences
are not large in view of those due to variations in
technical parameters (Table E.3).
Concerning the influence of tube voltage for the
same target type and HVL
1
of the incident radiation,
the variations reported by Rosenstein et al. (1985)
are somewhat larger than those published by Wu
et al. (1991, 1994). Wu et al. varied tube voltage
and also included an additional layer of Lexan
(polycarbonate). This latter filtration will have an
effect similar to the introduction of a compression
plate.
The influence on c
G
of different anode/filter com-
binations producing the same HVL
1
is dependent
on the compressed breast thickness (Dance, 1990).
Figure E.2 shows the conversion coefficients c
G
as a
function of HVL
1
for a 5-cm-thick standard breast,
i.e., containing a central region consisting of 50 %
adipose tissue and 50 % glandular tissue by mass
using various anode/filter combinations. The differ-
ence for a specific anode/filter combination from the
average value amounts to 5 % at maximum. This is
HVL
1
/ mm Al
C
O
N
V
E
R
S
I
O
N
C
O
E
F
F
I
C
I
E
N
T
;
C
G
Figure E.2. Calculated conversion coefficients for mammary gland c
G
for a 5.0-cm-thick breast model with a central region consisting of
50 % adipose and 50 % glandular tissue as a function of the HVL
1
for various anode/filter combinations (EC, 1996b). The anode the
anode material is indicated by the chemical symbol, the filter thickness is given in mm and the filter material by the trailing symbol.
Table E.3. Differences in conversion coefficients c
G
due to differences in technical parameters in mammography.
Technical parameter Variation
Tube voltage at same HVL
1
and target material Within 10 % from the average (Rosenstein, 1984)
Tube voltage at same HVL
1
and anode/filter combination,
but different layers of Lexan filtration
Within 5 % from the average (Wu et al., 1991)
Presence or absence of a compression plate 34 % (3 mm PMMA, Dance, 1990)
Difference in anode/filter combinations at same HVL
1
Few % from the average between W, MoW alloy and Mo targets
(Rosenstein, 1984)
3 % for a 4.5-cm-thick breast, for Mo/Mo, W/Mo, W/Rh, W/Pd and W/Al
combinations (Dance, 1990)
7 % difference between Mo/Rh and Rh/Rh (Wu et al., 1994)
Within 5 % for a 5-cm-thick standard breast for various anode/filter
combinations (Jansen, 1994)
Relative spectral correction factors are given by Dance et al. (2000).
They range up to 1.061 0.036 for Rh/Rh
APPENDIX E
95
in agreement with the findings by the other invest-
igators (Table E.2).
In addition, Wu et al. (1991) report a dependence
of c
G
on tube voltage waveform amounting to 4 %
between one phase, two pulse, and constant poten-
tial generators. This finding is less relevant today as
modern mammography units are equipped with
high-frequency tube-voltage generators, which pro-
vide approximately constant potentials.
Alm Carlsson and Dance (1992) report variations
in dose up to 4 % across an 8-cm-thick breast for a
homogeneous x-ray field. This variation is due to a
lack of side scatter and does not include a heel effect.
It is concluded that the differences in results
obtained by various investigators are rather small,
i.e., <9 %for HVL
1
in excess of 0.35 mmAl. The most
widely used conversion coefficients are those of
Wu et al. (1991) and Stanton et al. (1984) included
in the ACR Protocol (ACR, 1992) and of Dance
(1990) recommended in the European Protocol (EC,
1996b). The source of c
G
values employed should be
indicated.
According to the ACR protocol, K
a,i
(or incident
exposure) is measured for a mammographic phan-
tom (equivalent to 4.5 cm compressed breast tissue
with a composition of 50 % adipose tissue and 50 %
glandular tissue by weight; for example, Radiation
Measurement Inc. Model 156 Mammographic
Phantom or equivalent). The conversion coefficients,
c
G
for a 45-mm-thick breast containing a central
region (Figure E.1) consisting of 50 % adipose
tissue and 50 % glandular tissue are given in
Table E.4 (after ACR, 1992). The ACR conversion
coefficients have been converted to SI units.
In the European protocol on dosimetry for mam-
mography, K
a,i
is measured for 5-cm-thick breasts or
a 4.5-cm-thick PMMA reference phantom can be
used to assess D
G
for a 5-cm-thick standard breast
(EC, 1996b). The remaining dimensions of such a
reference phantom are either rectangular in excess
of 15 cm 10 cm or semicircular with a radius
>10 cm (EC, 1996b). The attenuation in a reference
PMMA phantom is somewhat smaller than that in
the corresponding 5-cm-thick standard breast model
used for numerical dosimetry. Therefore, the conver-
sion coefficients fromK
a,i
to D
G
are somewhat larger
than those for the 5-cm-thick standard breast and
are given in Table E.5 (EC, 1996b).
E.4 CONVERSION COEFFICIENTS C
G
AS A
FUNCTION OF BREAST THICKNESS
Almost all authors provide c
G
as a function of
HVL
1
and compressed breast thickness. As an
Table E.4. Conversion coefficients c
G
for calculating D
G
for a 4.5-cm-thick breast (50 % adipose/50 % glandular tissue) from
K
a,i
(after ACR, 1992. Reprinted with permission from the American College of Radiology).
HVL
1
b
Mo/30 mm Mo target-filter combination X-ray peak tube voltage/kV W/Al
a
23 24 25 26 27 28 29 30 31 32 33
0.23 0.124
0.24 0.129 0.132
0.25 0.133 0.137 0.139
0.26 0.138 0.141 0.144 0.146
0.27 0.144 0.146 0.148 0.151 0.153
0.28 0.148 0.151 0.153 0.155 0.157 0.158
0.29 0.154 0.156 0.158 0.161 0.162 0.163 0.164
0.30 0.159 0.161 0.163 0.165 0.166 0.168 0.169 0.170 0.194
0.31 0.164 0.166 0.168 0.170 0.171 0.172 0.173 0.174 0.176 0.200
0.32 0.169 0.171 0.172 0.174 0.176 0.177 0.178 0.180 0.181 0.182 0.182 0.205
0.33 0.174 0.176 0.177 0.179 0.180 0.181 0.182 0.185 0.186 0.187 0.187 0.211
0.34 0.179 0.181 0.182 0.184 0.185 0.186 0.187 0.189 0.190 0.192 0.192 0.217
0.35 0.186 0.187 0.189 0.190 0.192 0.193 0.194 0.195 0.196 0.196 0.221
0.36 0.192 0.194 0.195 0.196 0.197 0.198 0.200 0.201 0.201 0.227
0.37 0.198 0.200 0.201 0.202 0.203 0.203 0.204 0.205 0.233
0.38 0.204 0.205 0.206 0.208 0.208 0.209 0.210 0.237
0.39 0.210 0.211 0.212 0.212 0.213 0.214 0.243
0.40 0.216 0.217 0.218 0.219 0.219 0.247
0.41 0.221 0.222 0.223 0.223 0.252
0.42 0.228 0.228 0.257
0.43 0.233 0.262
0.44 0.267
0.45 0.271
a
W/Al target-filter combination, peak tube voltage 45 kV, filter thickness 1.6 mm.
b
HVL
1
in mm Al.
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
96
example, Table E.6 shows c
G
values as a function of
thickness of standard breasts, i.e., mathematical
models having a superficial layer of 0.5 cm adipose
tissue and a central region of 50 %adipose tissue and
50 % glandular tissue by mass. Application of these
conversion coefficients implies that the fraction of
glandular tissue is independent of compressed
breast thickness. This, however, is not the case in
practice (Geise and Palchevsky, 1996; Klein et al.,
1997; Young et al., 1998).
E.5 CONVERSION COEFFICIENTS C
G
FOR
ASSESSMENT OF D
G
TAKING BREAST
COMPOSITION INTO ACCOUNT
Conversion coefficients c
G
for fractions of glandu-
lar tissue in the central region of the compressed
breast differing from 50 % have been presented by
various investigators (Rosenstein et al., 1985; Wu et
al., 1991, 1994; Jansen et al., 1994; Klein et al., 1997;
Dance et al., 2000). Figure E.3 shows ratios of con-
version coefficients c
G
for breasts with glandularit-
ies different from 50 % for a HVL
1
of 0.35 mm Al.
A method for the assessment of the fraction
of glandular tissue has been presented by Klein
et al. (1997). Measurements were made of the
tube-current exposure-time product for radiographs
of phantoms of various thickness having central
regions with compositions of 0, 50, and 100 %
glandular tissue. By comparing the tube-current
exposure-time product measured for the phantoms
of different thickness and composition with that
determined during actual patient examinations at
known compressed breast thickness the fraction of
glandular tissue for a patient can be derived.
Klein et al. (1997) determined the fractions of
glandular tissue in German women. They found
that the average composition ranged from 75 % at a
compressed breast thickness of 25 mm through 40 %
at 50 mm breast thickness to an approximately con-
stant level of 20 % for breast thicknesses in excess of
70 mm. Similar work has been done by Geise and
Palchevsky (1996) and by Young et al. (1998) with
excellent and reasonable agreement with the results
of Klein et al. (1997), respectively.
An alternative method to take the fraction of
glandular tissue into account for assessment of D
G
has been presented by Dance et al. (2000). Dance
et al. (2000) propose to extend Eq. (E.1) to
D
G
K
a;i
c
G
cs, E:2
where c corrects for any difference in breast composi-
tion from 50 % glandular tissue, and the factor s for
any difference from the c
G
tabulation by Dance
(1990) due to the use of a different x-ray spectrum.
Correction factors c are tabulated against HVL
1
and
breast thickness. They are available as a function of
glandularity and for breasts of typical glandularity
at a given breast thickness for women in the age
groups 4049 years and 5064 years.
Table E.6. The conversion coefficient c
G
that relates K
a,i
to D
G
for standard breast phantoms (Reproduced from
Dance, 1990, with permission from IOP, UK).
HVL
1
/
mm Al
c
G
for breast thicknesses of
2 cm 3 cm 4 cm 4.5 cm 5 cm 6 cm 7 cm 8 cm
0.30 0.390 0.274 0.207 0.183 0.164 0.135 0.114 0.098
0.35 0.433 0.309 0.235 0.208 0.187 0.154 0.130 0.112
0.40 0.473 0.342 0.261 0.232 0.209 0.172 0.145 0.126
0.45 0.509 0.374 0.289 0.258 0.232 0.192 0.163 0.140
0.50 0.543 0.406 0.318 0.285 0.258 0.214 0.177 0.154
0.55 0.573 0.437 0.346 0.311 0.287 0.236 0.202 0.175
0.60 0.587 0.466 0.374 0.399 0.310 0.261 0.224 0.195
0.65 0.622 0.491 0.399 0.363 0.332 0.282 0.244 0.212
2 4 6 8
Breast thickness (cm)
0.6
0.8
1.0
1.2
1.4
R
a
t
i
o
o
f
c
o
n
v
e
r
s
i
o
n
f
a
c
t
o
r
s
0% glandularity
100% glandularity
Figure E.3. Ratios of conversion coefficients c
G
for breasts with
glandularities different from 50 % for a HVL
1
of 0.35 mm Al. The
upper curve is the ratio for 050 % and the lower curve is the ratio
for 10050 %. In each case the solid line corresponds to the data of
[Dance et al. (2000), reproduced with permission from IOP, UK],
the open circles to the data of Wu et al. (1991), and the closed
circles to the data of Klein et al. (1997).
Table E.5. Conversion coefficients for calculating D
G
for a
5.0 cmstandard breast fromK
a,i
for 4.5-cm-thick standard
PMMA phantom (EC, 1996b).
HVL
1
/mm Al c
G
0.30 0.177
0.35 0.202
0.40 0.223
0.45 0.248
0.50 0.276
0.55 0.304
0.60 0.326
0.65 0.349
APPENDIX E
97
Journal of the ICRU Vol 5 No 2 (2005) Report 74 doi:10.1093/jicru/ndi034
Oxford University Press
APPENDIX F: PCXMCA PC-BASED MONTE CARLO
PROGRAM FOR CALCULATING PATIENT DOSES
IN MEDICAL X-RAY EXAMINATIONS
F.1 INTRODUCTION
PCXMC is a PC-based Monte Carlo program for
calculating organ doses for patients undergoing
medical x-ray examination (radiography and fluoro-
scopy). The program uses the mathematical phan-
toms of Cristy (1980) and computes organ doses for
patients of different ages andsizes infreelyadjustable
x-ray projections and other examination conditions
used in radiology. The program has the unique fea-
ture of phantomscaling by changing weight or height.
The program runs on a PC operated under Win-
dows 95/98/NT, for example. The Monte Carlo simu-
lation time depends on the accuracy required and on
the speed of the PC. It takes typically from a few
seconds to a few minutes for a PC with a 200 MHz
Pentium processor. The same Monte Carlo data can
be used for calculating doses for any x-ray spectrum
of interest when the other conditions of the exam-
ination remain unchanged. In this case the calcula-
tion is fast because it does not involve any further
Monte Carlo simulation.
PCXMC has been developed by STUK, the
Radiation and Nuclear Safety Authority of Finland,
(Tapiovaara et al., 1997). A brief summary of the
program is given below. More information and a
free demonstration version of the program are avail-
able from the web page of STUK (http://www.stuk.fi/
pcxmc).
F.2 DOSE QUANTITIES IN PCXMC
All organ doses calculated by PCXMC are given in
proportion to the patient K
a,i
. Data on K
a,i
. (or,
alternatively, P
KA
) must be supplied by the user of
the program.
PCXMC calculates doses in the organs shown
below.
All doses in PCXMC are expressed in mGy.
Ovaries Oesophagus Pancreas
Testes Breasts Small intestine
Active BM Urinary bladder Spleen
Lungs Bone surface
a
Thymus
Colon
b
Skin Uterus
Stomach Adrenals Muscle
c
Liver Brain Gall bladder
Thyroid Kidneys Heart
a
PCXMC approximates the bone surface dose by using the dose to
the skeleton (excluding active BM).
b
The colon dose is calculated as D
colon
0.57 D
upper large intestine

0.43 D
lower large intestine
(see ICRP, 1995).
c
In PCXMC, this dose is calculated as the average dose to the
whole phantom, but excluding the other organs and tissues given
in this table.
Table F.1. Principal dimensions of the mathematical phantoms (Cristy 1980). The arms of a phantomcan be included at the
sides of the trunk or be removed.
Weight/
kg
Total height/cm Trunk height/cm Trunk thickness/cm Trunk width
a
/cm Trunk width
b
/cm Leg length/cm
Newborn 3.51 51.5 21.6 9.8 10.9 12.7 16.8
1 year old 9.36 75.0 30.7 13.0 15.1 17.6 26.5
5 years old 19.1 109.0 40.8 15.0 19.6 22.9 48.0
10 years old 32.1 138.6 50.8 16.8 23.8 27.8 66.0
15 years old 54.5 164.0 63.1 19.6 29.7 34.5 78.0
Adult 71.1 174.0 70.0 20.0 34.4 40.0 80.0
a
Excluding arms.
b
Including arms.
International Commission on Radiation Units and Measurements 2005
F.3 MATHEMATICAL PHANTOMS
The phantoms used in PCXMC are computa-
tional hermaphrodite phantoms representing
humans of various ages, i.e., newborn, 1, 5, 10,
15-year-old, and adult patients. The phantoms
include mathematical expressions describing vari-
ous organs and body parts. These phantoms have
been specified by Cristy (1980), but a few modifica-
tions, explained below, have been made. The prin-
cipal body dimensions of the phantoms are given in
Table F.1.
The origin of the coordinate system is located at
the middle of the base of the trunk of each phan-
tom. The z-axis points upwards, the x-axis to the
left-hand side of the phantom, and the phantom
looks in the negative y-direction.
Cristys phantom models have been developed for
calculating doses from internal radiation sources. To
simulate the conditions in radiology, the following
changes, closely following Jones and Wall (1985) and
Zankl et al. (1992), have been made to the Cristy
phantoms.
The esophagus has been added. The dimensions
and location of the esophagus are modelled to
resemble those of the male phantom model of
Zankl et al. (1992) and are given for the phantoms
of various ages by Tapiovaara et al. (1997).
The anterior half of the neck of the phantom has
been modified from the Cristy phantom model,
which has no jaw (Kramer et al., 1982). This
modification makes the depths of the thyroid
and esophagus better resemble those in real
patients. The dimensions of the anterior neck
region are given by Tapiovaara et al. (1997).
The arms of the phantoms can be present or
removed. The latter geometry may simulate the
real situation better, e.g., for lateral projections.
The removal of the arms is accomplished by cut-
ting the trunk region with two planes, which are
parallel to the yz plane, and located at the
maximum dimension of the rib cage in the direc-
tion of the x-axis, but increased by the thickness
of the skin.
The breast material is taken to be a 50:50 mix-
ture of fat and water.
PCXMC allows further modification of the phan-
toms by letting the user change their weight, M, or
height, h. Using these body size measurements, the
program calculates scaling factors
s
z
h=h
0
, F:1
s
xy
h
o
M=hM
o

0:5
, F:2
where s
z
is the scaling factor in the direction of
the z-axis (phantom height), s
xy
the scaling factor
in the directions of the x and y-axes (phantom
width and thickness, respectively), and h
o
and M
o
are the height and weight of the basic phantom
(Table F.1). All dimensions of the phantoms are
then multiplied by these scaling factors, and the
organ masses are changed accordingly. This opera-
tion allows the shape of the phantoms to be modified
to resemble that of the actual patient more closely. It
should be noted, however, that all measures in a
given direction (i.e., vertical or horizontal) are
multiplied by the same scaling factor and phantom
variability due to, for example, variability in the
amount of fat tissue cannot be simulated by this
method.
The elemental composition of the tissues of the
phantoms is shown in Table F.2. The number of
elements included in PCXMC is reduced compared
with that employed by the NRPB (Jones and Wall,
1985). The elements Na, Mg, P, S, and Cl are
grouped together and treated as phosphorus. All
elements of atomic numbers Z19 are grouped
together and treated as calcium (Z 20).
F.4 THE MONTE CARLO RADIATION
TRANSPORT METHOD
The Monte Carlo simulation has been formulated
in PCXMC in a standard manner. Photons are
assumed to be emitted from an isotropic point source
into the solid angle specified by the focus-to-surface-
distance and the x-ray field dimensions. The prim-
ary photons are followed while they interact with the
phantomaccording to the probability distributions of
the physical processes that they may undergo, i.e.,
photoelectric absorption, coherent scattering, or
Table F.2. Elemental composition of the phantom tissues as used in PCXMC (percent by weight).
Density/g cm
3
H C N O P Ca
Skeleton 1.49 7.04 22.79 3.87 48.56 7.68 10.06
Active BM 0.99 10.18 47.48 2.18 39.70 0.29 0.17
Breast tissue 0.975 11.70 38.04 0.00 50.26 0.00 0.00
Lung tissue 0.296 10.21 10.01 2.80 75.96 0.78 0.24
Other soft tissues 0.99 10.47 23.02 2.34 63.21 0.39 0.58
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
100
incoherent scattering. Other interactions are not
considered in PCXMC, because the maximum
photon energy is limited to 150 keV. The range of
secondary electrons in soft tissue is only a fraction of
a millimeter, and the energy of the secondary elec-
trons is approximated to be absorbed at the site of
the photon interaction (except in calculating the BM
dose, see below). A large number of individual
photon histories are generated, and estimates of
the mean values of the energy depositions in the
various organs in the phantom are used for calculat-
ing the dose in these organs.
Pseudo-random numbers are used for sampling
the initial photon direction, distance between inter-
actions, type of interacting atom, type of interaction,
scattering angle, and energy. The cross sections for
the photoelectric interaction, coherent scattering,
and incoherent scattering have been taken from
Storm and Israel (1970) and the atomic form factors
and incoherent scattering functions from Hubbell
et al. (1975).
The bones of the mathematical phantoms are mod-
elled as a homogeneous mixture of mineral bone,
active BM, and other organic constituents of the
skeleton. The overall composition of the skeleton is
approximated as being constant over all bones in the
body and over all phantoms representing patients of
various ages (Table F.2). However, the amount of
active BM is varied from one part of the skeleton to
another and is different for phantoms representing
different ages (Cristy, 1980). In reality, active BM is
located in small cavities in the trabecular bone, and
this must be taken into account when calculating the
dose to the active BM. PCXMC calculates the kerma
in both components of the skeleton, the active BM
and the rest of skeletal material, by dividing the
absorbed energy in the whole skeleton into two
parts, applying the method of Rosenstein (1976). In
PCXMC, however, the actual energy spectrum at the
interaction site is used instead of the incident photon
spectrum used by Rosenstein. The influence of the
small cavity size on the dose in the active BM is
considered by applying a photon energy dependent
kerma-to-dose-conversion coefficient (Kerr and Eck-
erman, 1985), which increases the active BM dose by
a few percentage compared with the kerma. The size
of the BM cavities may vary depending on the age or
the anatomical part of the skeleton (Kerr and Eck-
erman, 1985), but this has not been taken into
account in PCXMC.
PCXMC calculates the organ doses for mono-
chromatic photons of 10150 keV in energy steps of
10 keV (or up to less than the maximum 150 keV,
if the user so specifies) in 10 batches at each
Dose in active bone marrow
0
50
100
150
200
250
300
0
h
a
1
2
0
1
5
c
p
1
2
0
0
a
a
1
2
0
1
h
a
1
2
0
0
c
a
1
2
0
1
0
a
a
1
2
0
0
p
p
1
2
0
1
0
c
a
1
2
0
1
c
a
1
2
0
5
c
a
1
2
0
1
a
a
1
2
0
1
5
c
a
1
2
0
1
5
a
a
1
2
0
5
c
r
1
2
0
0
h
a
6
0
5
h
a
1
2
0
0
a
a
6
0
1
h
a
6
0
0
c
a
6
0
1
5
c
p
6
0
0
p
p
6
0
1
0
h
a
1
2
0
1
c
a
6
0
1
0
c
a
6
0
1
5
h
a
1
2
0
5
c
a
6
0
1
5
c
a
6
0
1
0
a
a
6
0
1
5
h
p
1
2
0
5
a
a
6
0
5
c
r
6
0
1
a
a
6
0
5
h
a
6
0
1
5
a
a
6
0
1
0
h
a
6
0
1
5
h
a
6
0
1
5
h
p
6
0
Examination condition
m
G
y
/
G
y
PCXMC
NRPB
Figure F.1. A comparison of the conversion factors calculated by PCXMC with the data of Hart et al. (1996b) according to (Tapiovaara
et al., 1997). Conversion coefficient from air kerma to dose in the active BM. All doses correspond to a K
a,i
value of 1 Gy. In the examina-
tion condition, the first digits show the patients age, the letters show the examination (a, abdomen; c, chest; h, head; p, pelvis) and
projection (a, AP; p, PA; r, right lateral), and the last digits show the x-ray spectrum (60, 60 kV, total filtration 3 mm Al; 120, 120 kV, total
filtration 3 mm Al and 0.2 mm Cu, 17

x-ray tube anode angle). The uncertainty bars shown correspond to two standard deviations of
the data.
APPENDIX F
101
energy. The final estimate of the absorption at each
energy is obtained as the average of these batches,
and the statistical uncertainty is estimated from
their standard deviation. The doses and their stat-
istical uncertainties for a practical x-ray spectrum of
interest are calculated afterwards by another mod-
ule included in the program. The same Monte Carlo
data can, therefore, be used for calculating doses for
any spectrum of interest. The calculation is fast
because it does not involve any further Monte Carlo
simulation.
The x-ray spectra are calculated according to the
theory of Birch and Marshall (1979) and specified in
terms of the x-ray tube voltage (kV), the angle of the
tungsten target of the x-ray tube, and filtration. In
the present version of the program, the user can
simultaneously define two filters of arbitrary atomic
numbers and thickness. The filter data are from the
compiled x-ray interaction data of McMaster et al.
(1969). Air kerma is calculated from photon fluence
data using the conversion coefficients in ICRU
Report 46 (1992c).
F.5 COMPARISON WITH OTHER DATA
The data calculated by PCXMC are most directly
comparable with the organ dose-conversion coeffi-
cients calculated at the NRPB by Jones and Wall
(1985) and Hart et al. (1994b, 1996b), because they
are based on the same phantom models (Cristy,
1980). In general, the agreement between the results
of PCXMC and the NRPB data is good (e.g.,
Figure F.1). The differences typically fall within
the statistical uncertainty and are generally <10
%. The remaining deviations can be explained by
the differences in the phantoms. An exception to
the good agreement is the calculations for esopha-
gus, where differences in the phantom model lead to
larger deviations (by up to a factor of 2, Tapiovaara
et al., 1997).
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
102
Journal of the ICRU Vol 5 No 2 (2005) Report 74 doi:10.1093/jicru/ndi035
Oxford University Press
REFERENCES
AAPM. (1992) American Association of Physicists in Medi-
cine. Recommendations on Performance Characteristics
of Diagnostic Exposure Meters. AAPM Report No 35
(New York: American Institute of Physics) (Also pub-
lished in Med. Phys. (1992) 19, 231241).
ACR. (1992) American College of Radiology. Mammo-
graphy Quality Control (Merrifield, VA: American Col-
lege of Radiology).
Adrian Committee. (1960) Radiological Hazards to
Patients. Second Report to the Committee (London:
Her Majestys Stationary Office).
Allisy, A. (1995) An early proposal for the measurement of
Rontgen rays (18961897). ICRU News 1, 58.
Alm Carlsson, G. and Dance, D. R. (1992) Breast absorbed
doses in mammography: evaluation of experimental and
theoretical approaches. Radiat. Prot. Dosim. 43, 197-
200.
Andersen, L. W. and Rosenstein, M. (1985) Computer Pro-
gram for Absorbed Dose to the Breasts in Mammo-
graphy. HHS Publication FDA 85-8243 (Rockville, MD:
Center for Devices and Radiological Health).
Arai, T., Nakano, T., Fukuhisa, K., Tsunematsu, R., Mas-
ubuchi, K., Yamauchi, K., Hamada, T., Fukuda, T.,
Noguchi, H. and Murata, M. (1991) Second cancer
after radiation therapy for cancer of the uterine cervix.
Cancer 67, 398405.
Arnal, M. L. and Pychlau, H. (1961) Die Strahlenbelas-
tung des Patienten bei rontgendiagnostischen
Untersuchungen. Fortschr. Geb. Rontgenstr. 95,
323335.
Baorong, Y., Kramer, H. M., Selbach, H.-J. and Lange, B.
(2000) Experimental determination of practical peak
voltage. Br. J. Radiol. 73, 641649.
Beck, T. J. and Rosenstein, M. (1979) Quantification of
Current Practice in Pediatric Rontgenography for
Organ Dose Calculations. HEW Publication FDA 79-
8078 (Rockville, MD: Bureau of Radiological Health).
Bednarek, D. R. and Rudin, S. (2000) Comparison of
two dose-area-product ionization chambers with
different conductive surface coating for over-table and
under-table tube configurations. Health Phys. 78,
316321.
Berger, M. J. and Hubbell, J. H. (1987) XCOM: Photon
Cross Sections on a Personal Computer (Gaithersburg,
MD: National Bureau of Standards).
BIPM, IEC, ISO and OIML. (1984) International Bureau
of Weights and Measures, International Electrotech-
nical Commission, International Organization for
Standardization and International Organization of
Legal Metrology. International Vocabulary of Basic
and General Terms in Metrology (Geneva: International
Organization for Standardization).
Birch, R. and Marshall, M. (1979) Computation of brems-
strahlung X-ray spectra and comparison with spectra
measured with a Ge (Li) detector. Phys. Med. Biol. 24,
505517.
Birch, R., Marshall, M. and Ardran, G. M. (1979) Cata-
logue of Spectral Data of Diagnostic X Rays. Hospital
Physicists Association, SRS-30 (York: Institute of Phys-
ics and Engineering in Medicine).
BJR. (1983) British Journal of Radiology. Central axis
depth dose data for use in radiotherapy. Br. J. Radiol.
Supplement 17.
Boag, J. W. (1987) Ionization chambers. In: The Dosimetry
of Ionizing Radiation, vol. II, Kase, K. R., Bja rngard, B.
G. and Attix, F. H., Eds. (Orlando: Academic Press) pp.
169243.
Boice, J. D. Jr, Day, N. E., Andertsen, A., Brinton, L. A.,
Brown, R., Choi, N. W., Clarke, E. A., Coleman, M. P.,
Curtis, R. E., Flannery, J. T., Hakama, M., Hakulinen,
T., Howe, G. R., Jensen, O. M., Kleinerman, R. A., Mag-
nin, D., Magnus, K., Makela, K., Malker, B., Miller, A.
H., Nelson, N., Patterson, C. C., Pettersson, F., Pom-
pekirn, V., Primiczakelj, M., Prior, P., Ravnihar, B.,
Skeet, R. G., Skjerven, J. E., Smith, P. G., Sok, M.,
Spengler, R. F., Storm, H. H., Stovall, M., Tomkins, G.
W. O. and Wall, C. (1985) Second cancers following
radiation treatment for cervical cancers: An interna-
tional collaboration among cancer registries. J. Natl
Cancer Inst. 74, 955975.
Boice J. D. Jr, Engholm, G., Kleinerman, R., Blettner,
M., Stovall, M., Lisco, H., Moloney, W. C., Austin, D.
F., Bosch, A., Cookfair, D. L., Krementz, E. T.,
Latourette, H. B., Merrill, J. A., Peters, L. J., Schulz,
M. D., Storm, H. H., Bjorkholm, E., Pettersson, F.,
Bell, C. M. I., Coleman, M. P., Fraser, P., Neal, F. E.,
Prior, P., Choi, N. W., Hislop, T. G., Koch, M.,
Kreiger, N., Robb, D., Robson, D., Thomson, D. H.,
Lochmuller, H., Vonfournier, D., Frischkorn, R.,
Kjorstad, K. E., Rimpela, A., Pejovic, M. H., Kirn,
V. P., Stankusova, H., Berrino, F., Sigurdsson, K.,
Hutchison, G. B. and Macmahon, B. (1988) Radiation
se and second cancer risk in patient treated for
cancer of the cervix. Radiat. Res. 116, 355.
Bos, A. J. J., Piters, T. M., Ros, J. M. G. and Delgado, A.
(1993) An intercomparison of glow curve analysis com-
puter programs 1. Synthetic glow curves. Radiat. Prot.
Dosim. 47, 473477.
International Commission on Radiation Units and Measurements 2005
Bos, A. J. J., Piters, T. M., Ros, J. M. G. and Delgado, A.
(1994) An intercomparison of glow curve analysis com-
puter programs 2. Measured glow curves. Radiat. Prot.
Dosim. 51, 257264.
Braun, R., Hase, H. and Ku stner, H. (1928) U

ber die in der


Diagnostik verabfolgten Dosen in R-Einheiten. Forts-
chr. D. Rontgenstr. 38, 385398.
Brenner, D. J. and Amols, H. I. (1989) Enhanced risk from
low-energy screen-film mammography X-rays. Br. J.
Radiol. 62, 910914.
Brenner, D. J., Doll, R., Goodhead, D. T., Hall, E. J., Land,
C. E., Little, J. B., Lubin, J. H., Preston, D. L., Puskin, J.
S., Ron, E., Sachs, R. K., Sametl, J. M., Setlow, R. B. and
Zaider, M. (2003) Cancer risk attributable to low doses
of ionizing radiation; assessing what we really know.
Proc. Natl Acad. Sci USA 100, 1376113766.
Burkhardt, R. (1984) Nationwide Evaluation of X-Ray
Trends (NEXT): Eight Years of Data (19741981)
(Springfield, VA: National Technical Information Ser-
vice).
Cameron, J. R. (1971) A proposed unit for patient radi-
ation exposure from diagnostic x-rays. Health Phys. 21,
879880.
Carlsson, C. (1963) Determination of integral absorbed
dose from exposure measurements. Acta Radiol. Ther.
Phys. Biol. 3, 384408.
Carlsson, C. A. (1993) Differences in reported backscatter
factors for low-energy x rays: a literature study. Phys.
Med. Biol. 38, 521531.
Carlsson, C. A. and Alm Carlsson, G., (1990) Dosimetry in
diagnostic radiology and computed tomography. In: The
Dosimetry of Ionizing Radiation, vol. III, Kase, K. R.,
Bja rngard, B. E., Attix, F. H., Eds. (Orlando: Academic
Press) pp. 163257.
CEC. (1990) Quality Criteria for Diagnostic Radiographic
Images, 2nd edition. Working Document XII/173/90
(Brussels: Commission of the European Communities).
Chan, H.-P. and Doi, K. (1981) Monte Carlo simulation
studies of backscatter factors in mammography. Radi-
ology 139, 195199.
Chen, W., Poston, J. W. and Warner, G. G. (1978) An
Evaluation of the Distribution of Absorbed Dose in
Child Phantoms Exposed to Diagnostic Medical X
Rays. ORNL-TM-5933 (Oak Ridge, TN: Oak Ridge
National Laboratory).
Cohen, M., Jones, D. E. A. and Green, D., Eds. (1972)
Central axis depth dose data for use in radiotherapy.
Br. J. Radiol. Supplement 11.
Conway, B. J. (1994) Nationwide Evaluation of X-Ray
Trends (NEXT). Summary of 1990 Computerized
Tomography Survey and 1991 Fluoroscopy Survey.
CRCPD Publication 94-2, CRCPD Committee (H-4)
(Frankfort, KY: Conference of Radiation Control Pro-
gram Directors).
Conway, B. J., Butler, P. F., Duff, J. E., Fewell, T. R.,
Gross, R. E., Jennings, R. J., Koustenis, G. H., McCro-
han, L. J., Rueter, F. G. and Showalter, C. K. (1984)
Beam quality independent attenuation phantom for
estimating patient exposure from X-Ray automatic
exposure controlled chest examinations. Med. Phys.
11, 827832.
Conway, B. J., Duff, J. E., Fewell, T. R., Jennings, R. J.
Rothenverg, L. N. and Fleischman, R. C. (1990)
A patient equivalent attenuation phantom for estimat-
ing patient exposures from automatic exposure con-
trolled X-ray examinations of the abdomen and lumbo-
sacral spine. Med. Phys. 17, 448453.
CRCPD. (2002) Conference on Radiation Control Program
Directors. Nationwide Evaluation of X-ray Trends
(NEXT). Tabulation and Graphical Summary of 1996
Fluoroscopy Survey. CRCPD Publication E-02-4 (USA:
Conference of Radiation Control Program Directors
Inc.).
Cristy, M. (1980) Mathematical Phantoms Representing
Children of Various Ages for Use in Estimates of
Internal Dose. NUREG/CR-1159, ORNL/NUREG/TM-
367 (Oak Ridge, TN: Oak Ridge National Laboratory).
Cristy, M. (1981) Active bone marrow distribution as a
function of age in humans. Phys. Med. Biol. 26,
389400.
Dance, D. R. (1990) Monte Carlo calculation of conversion
factors for the estimation of mean glandular breast
dose. Phys. Med. Biol. 35, 12111219.
Dance, D. R., Skinner, C. L., Young, K. C., Beckett, J. R.
and Kotre, C. J. (2000) Additional factors for the
estimation of mean glandular breast dose using the
UK mammography dosimetry protocol. Phys. Med.
Biol. 45, 32253240.
De Sousa, M. C., Aubert, B. and Ricard, M. (2000)
Evaluation of physical performance of a scintillation
dosemeter for patient dosimetry in diagnostic radi-
ology. Br. J. Radiol. 73, 12971305.
DeWerd, L. A., Micka, J. A., Laird, R. W., Pearson, D. W.,
OBrien, M. and Lamperti, P. (2002) The effect on cal-
ibration and measurement with mammographic ioniza-
tion chambers. Med. Phys. 29, 26492654.
Doi, K. and Chan, H. P. (1980) Evaluation of absorbed
dose in mammography: Monte Carlo simulation
studies. Radiology 135, 199208.
Doi, K., Holje, G., Loo, L.-N., Chan, H. P., Sandrik, J.
M., Jennings, R. J. and Wagner, R. F. (1982) MTFs
and Wiener Spectra of Radiographic Screen-Film Sys-
tems. U.S. Department of Health and Human Ser-
vices Publication Food and Drug Administration
82-8187.
Doi, K., Kodera, Y., Loo, L.-N., Chan, H. P. and Higashida,
Y. (1986) MTFs and Wiener Spectra of Radiographic
Screen-Film Systems, vol. 2. U.S. Department of Health
and Human Services Publication Food and Drug
Administration 86-8257.
Drexler, G., Panzer, W., Widenmann, L., Williams, G.
and Zankl, M. (1990) The Calculation of Dose from
External Photon Exposures Using Reference Human
Phantoms and Monte Carlo Methods. Part III: Organ
Doses in X-Ray Diagnosis. Revised and Amended,
GSF-Bericht 11/90 (Neuherberg: GSF-National
Research Centre for Environment and Health, D-
85764, ISSN 0721-1694).
Drexler, G., Panzer, W., Stieve, F.-E., Widenmann, L. and
Zankl, M. (1993) Die Bestimmung von Organdosen in
der Rontgendiagnostik. Zweite, Verbesserte Auflage
(Berlin: H. Hoffmann Verlag).
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
104
Dutreix, A. and Bridier, A. (1985) Dosimetry for external
beams of photon and electron radiation. In: The Dosi-
metry of Ionizing Radiation, vol. I, Kase, K. R., Bja rn-
gard, B. E. and Attix, F. H., Eds. (Orlando, FL:
Academic Press) pp. 164228.
EC. (1996a) European Commission. The 1991 CEC Trial
on Quality Criteria for Diagnostic Radiographic
Images: Detailed Results and Findings. Report EUR
16635 (Luxembourg: Office for Official Publications of
the European Communities).
EC. (1996b) European Commission. European Protocol on
Dosimetry in Mammography. Report EUR 16263 (Lux-
embourg: Office for Official Publications of the Euro-
pean Communities).
EC. (1996c) European Commission. European Guidelines
on Quality Criteria for Diagnostic Radiographic
Images. Report EUR 16260 (Luxembourg: Office for
Official Publications of the European Communities).
EC. (1996d) European Commission. European Guidelines
on Quality Criteria for Diagnostic Radiographic Images
in Paediatrics. Report EUR 16261 (Luxembourg: Office
for Official Publications of the European Communities).
EC. (1997) European Commission. Council directive of
June 30, 1997 (97/43/Euratom) on health protection of
individuals against the dangers of ionizing radiation in
relation to medical exposure. Official J. Eur. Commun.
No. L180/22.
EC. (2000) European Commission. European Guidelines
on Quality Criteria for Computed Tomography. Report
EUR 16262 (Luxembourg: Office for Official Publica-
tions of the European Communities).
Ellis, R. E. (1961) The distribution of active bone marrow
in the adult. Phys. Med. Biol. 5, 255259.
Ellis, R. E., Healy, M. J. R., Shleien, B. and Tucker, T.
(1975) A System for Estimation of Mean Active Bone
Marrow Dose. HEW Publication FDA 76-8015 (Rock-
ville, MD: Bureau of Radiological Health).
FDA. (1997) Food and Drug Administration. Computed
Tomography (CT) equipment. Section 1020.33, Title 21
Code of Federal Regulations, Chapter J, Radiological
Health, 1997 edition (Rockville, MD: U.S. Food and
Drug Administration).
Fewell, T. R. and Shuping, R. E. (1978) Handbook of Mam-
mographic X-Ray Spectra. HEW Publication FDA 79-
8071 (Rockville, MD: Bureau of Radiological Health).
Feller, W. (1968) An Introduction to Probability Theory
and Its Applications, vol. I (New York: John Wiley and
Sons).
Fill, U. A. and Regulla, D. F. (1998) Advanced LiF techno-
logy for the assessment of patient exposure in diagnostic
radiology. Radiat. Prot. Dosim. 80, 225229.
Gard, M. F. (1996) Measurement of x-ray tube accelerat-
ing potential by differential beam filtration. IEEE
Trans. Instrum. Meth. 45, 745749.
Geise, R. A. and Palchevsky, A. (1996) Composition of
mammographic phantom materials. Radiology 198,
347350.
Geise, R. A., Schueler, B. A., Lien, W. and Jones, S. C.
(1997) Suitability of laser stimulated TLD arrays as
patient dose monitors in high dose x-ray imaging.
Med. Phys. 24, 16431646.
Geleijns, J. and Wondergem, J. (2005) X-ray imaging and
the skin: radiation biology, patient dosimetry and
observed effects. Radiat. Prot. Dosim. 114, 121125.
Geleijns, J., Broerse, J. J., Julius, H. W., Vrooman, H. A.,
Zoetelief, J., Zweers, D. and Schultze Kool, L. J. (1992)
Dosimetric aspects of a scanning multiple beam equal-
ization system for chest radiography (AMBER). Radiat.
Prot. Dosim. 43, 155159.
Geleijns, J., Van Unnik, J. G., Zoetelief, J., Zweers, D. and
Broerse, J. J. (1994) Comparison of two methods for
assessing patient dose from computed tomography,
Brit. J. Radiol. 67, 360365.
Geleijns, J., Broerse, J. J., Chandie Shaw, M. P., Schultz,
F. W., Teeuwisse, W., Van Unnik, J. G. and Zoetelief, J.
(1997) Patient dose due to colon examination: dose
assessment and results from a survey in the Nether-
lands. Radiology 204, 553559.
Gentry, J. R. and DeWerd, L. A. (1996) TLD measure-
ments of in vivo mammographic exposures and the cal-
culated mean glandular dose across the United States.
Med. Phys. 23, 899903.
Giles, E. R. and Murphy, P. H. (2002) Measuring skin dose
with radiochromic dosimetry film in the cardiac cathet-
erization laboratory. Health Phys. 82, 875880.
Gilmore, B. J. and Cranley, K. (1990) Errors in the deter-
mination of total filtration of diagnostic x-ray tubes by
the HVL method. Phys. Med. Biol. 35, 9991015.
Gkanatsios, N. A., Huda, W., Peters, K. R. and
Freeman, J. A. (1997) Evaluation of an on-line patient
exposure meter in neuroradiology. Radiology 203,
837842.
Gosch, D. and Gursky, S. (1992) Describing the radiation
exposure of patients in diagnostic radiology on the basis
of absorbed energy. Radiat. Prot. Dosim. 43, 115117.
Gregan, A. C. M., Peach, D. and McHugo, J. M. (1998)
Patient dosimetry in hysterosalpingography: a compar-
ative study. Br. J. Radiol. 71, 10581061.
Grosswendt, B. (1984) Backscatter factors for x-rays gen-
erated at voltages between 10-100 kV. Phys. Med. Biol.
29, 579591.
Grosswendt, B. (1990) Dependence of the photon backs-
catter factor for water on source-to-phantom distance
and irradiation field size. Phys. Med. Biol. 35,
12331245.
Grosswendt, B. (1993) Dependence of the photon backs-
catter factor for water on irradiation field size and
source-to-phantom distance between 1.5 and 10 cm.
Phys. Med. Biol. 38, 305310.
Guibelalde, E., Van o, E., Gonzales, L., Prieto, C.,
Fernandez, J. M. and Ten, J. I. (2003) Practical aspects
for the evaluation of skin doses in interventional cardi-
ology using a new slow film. Br. J. Radiol. 76, 332336.
Haddadi, R. and Renaud, L. (1993) Projections et Condi-
tions Techniques en Usage en Angiocardiologie: Etude
Statistique, Rapport Technique (Montreal: Service de
Genie Biomedical, Institut de Cardiologie de Montreal).
Hall, E. J. (2000) Radiobiology for the Radiologist, 5th
edition. (New York: Lippincott Williams and Wilkins).
Hall, P. (2002) Health consequences after irradiation in
utero-human data. Risk of leukemia and mental
retardation. In: Proceedings of a scientific seminar
REFERENCES
105
held in Luxembourg on 5 November 2001 (Luxembourg:
Office for Offical Publications of the European Commu-
nities) pp. 3756.
Hall, P., Adami, H. O., Trichopoulos, D., Pedersen, N. L.,
Lagiou, P., Ekbom, A., Ingvar, M., Lundell, M. and
Granath, F. (2004) Effect of low doses of ionising radi-
ation in infancy on cognitive function in adulthood:
Swedish population based cohort study. Br. Med. J.
328, 1921.
Hammerstein, G. R., Miller, D. W., White, D. R., Master-
son, M. E., Woodard, H. Q. and Laughlin, J. S. (1979)
Absorbed radiation dose in mammography. Radiology
130, 485-491.
Harrison, R. M. (1982) Backscatter factors for diagnostic
radiology (14 mm Al HVL). Phys. Med. Biol. 27,
14651474.
Harrison, R. M., Walker, C. and Aukett, R. J. (1990)
Measurement of backscatter factors for low energy
radiotherapy (0.12.0 mm Al HVL) using thermolumin-
escence dosimetry. Phys. Med. Biol. 35, 12471254.
Hart, D., Jones, D. G. and Wall, B. F. (1994a) Estimation of
Effective Dose in Diagnostic Radiology from Entrance
Surface Dose and Dose-Area Product Measurements.
NRPB-R262 (London, UK: Her Majestys Stationary
Office).
Hart, D, Jones, D. G. and Wall, B. F. (1994b) Normal-
ised Organ Doses for Medical X-ray Examinations
Calculated Using Monte Carlo Techniques. NRPB-
SR262 (Chilton, UK: National Radiological Protection
Board).
Hart, D, Jones, D. G. and Wall, B. F. (1996a) Coefficients
for Estimating Effective Dose from Paediatric X-ray
Examinations. NRPB-R279 (London, UK: Her Majestys
Stationary Office).
Hart, D, Jones, D. G. and Wall, B. F. (1996b) Normalised
Organ Doses for Paediatric X-ray Examinations Calcu-
lated Using Monte Carlo Techniques. NRPB-SR279
(Chilton, UK: National Radiological Protection Board).
Hart, D., Hillier, M. C., Wall, B. F., Shrimpton, P. and
Bungay D. (1996c) Doses to Patients from Medical
X-ray Examinations in the UK 1995. Report NRPB-
R289 (London, UK: Her Majestys Stationary Office).
Havukainen, R. and Pirinen, M. (1993) Patient dose and
image quality in five X-ray examinations. Med. Phys.
20, 813817.
Horowitz, Y.S. (1984) Thermoluminescence and Thermo-
luminescent Dosimetry (CRC Press, Inc., Boca Raton,
Florida, USA).
Hubbell, J. H. and Seltzer S. M. (1995) Tables of X-ray
Mass Attenuation and Mass Energy-Absorption Coeffi-
cients from 1 keV to 20 MeV for Elements Z 1 to 92 and
48 Additional Substances of Dosimetric Interest.
NISTIR 5632 (Gaithersburg, MD: National Institute of
Standards and Technology).
Hubbell, J. H., Veigele, W. J., Briggs, E. A., Brown, R. T.,
Cromer, D. T. and Howerton, R. J. (1975) Atomic form
factors, incoherent scattering functions and photon
scattering cross-sections. J. Phys. Chem. Ref. Data 4,
471538.
Hwang, E., Gaxiola, E., Vlietstra, R. E., Brenner, A., Eber-
sole, D. and Browne K. (1998) Real-time measurement
of skin radiation during cardiac catheterization.
Cathet. Cardiovasc. Diagn. 43, 367370.
Hwang, J. M. L., Shoup, R. L. and Poston, J. W. (1976a)
Mathematical Description of a Newborn for Use in Dosi-
metry Calculations. ORNL-TM-5433 (Oak Ridge, TN:
Oak Ridge National Laboratory).
Hwang, J. M. L., Shoup, R. L., Warner, G. G. and Poston, J.
W. (1976b) Mathematical Descriptions of a One and Five
Year Old Child for Use in Dosimetry Calculations.
ORNL-TM-5293 (Oak Ridge, TN: Oak Ridge National
Laboratory).
IAEA. (1987) International Atomic Energy Agency.
Absorbed Dose Determination in Photon and Electron
Beams. IAEA Technical Reports Series 277 (Vienna:
International Atomic Energy Agency).
IAEA. (1994) International Atomic Energy Agency. Calib-
ration of Dosimeters used in Radiotherapy. IAEA Tech-
nical Report Series 374 (Vienna: International Atomic
Energy Agency).
IAEA. (1996a) International Atomic Energy Agency. Inter-
national Basic Safety Standards for Protection against
Ionizing Radiation and for the Safety of Radiation
Sources. IAEA Safety Series 115 (Vienna: International
Atomic Energy Agency).
IAEA. (1996b) International Atomic Energy Agency.
Review of Data and Methods Recommended in the Inter-
national Code of Practice. IAEA Technical Report Series
No. 227, Absorbed Dose Determination in Photon and
Electron Beams, IAEA-TECDOC-897 (Vienna: Interna-
tional Atomic Energy Agency).
IAEA. (2000) International Atomic Energy Agency. Calib-
ration of Radiation Protection Monitoring Instruments.
IAEA Safety Reports Series No. 16 (Vienna: Interna-
tional Atomic Energy Agency).
ICRP. (1969) International Commission on Radiological
Protection. Radiosensitivity and Spatial Distribution
of Dose. ICRP Publication 14 (London: Pergamon
Press).
ICRP. (1975) International Commission on Radiological
Protection. Reference Man: Anatomical, Physiological
and Metabolic Characteristics. ICRP Publication 23
(Oxford: Pergamon Press).
ICRP. (1987) International Commission on Radiological
Protection. Recommendations of the International Com-
mission on Radiological Protection. ICRP Publication
52. Ann. ICRP 17(4) (Oxford: Pergamon Press).
ICRP. (1991a) International Commission on Radiological
Protection. 1990 Recommendations of the International
Commission on Radiological Protection. ICRP Publica-
tion 60. Ann. ICRP 21(13) (Oxford: Pergamon Press).
ICRP. (1991b) International Commission on Radiological
Protection. The Biological Basis for Dose Limitation in
the Skin. ICRP Publication 59. Ann. ICRP 22(2)
(Oxford: Pergamon Press).
ICRP. (1995) International Commission on Radiological
Protection. Age Dependent Doses to Members of the Pub-
lic from Intake of Radionuclides: Part 4, Inhalation
Dose Coefficients. ICRP Publication 71. Ann. ICRP
25(34) (Oxford: Pergamon Press).
ICRP. (1996) International Commission on Radiological
Protection. Radiological Protection and Safety in Medi-
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
106
cine. ICRP Publication 73. Ann. ICRP 26(2) (Oxford:
Pergamon Press).
ICRP. (2000) International Commission on Radiological
Protection. Avoidance of Radiation Injuries from Inter-
ventional Procedures. ICRP Publication 85. Ann. ICRP
30(2) (Oxford: Pergamon Press).
ICRU. (1961) International Commission on Radiation
Units and Measurements. Report of the ICRU, NBS
Handbook 78 (Washington, DC: US Government Print-
ing Office).
ICRU. (1964) International Commission on Radiation
Units and Measurements. Physical Aspects of Irradi-
ation. ICRU Report 10b, NBS Handbook 85 (Washing-
ton, DC: US Government Printing Office).
ICRU. (1970) International Commission on Radiation
Units and Measurements. Radiation Dosimetry: X-
Rays Generated at Potentials of 5 to 150 kV. ICRU
Report 17 (Bethesda, MD: International Commission
on Radiation Units and Measurements).
ICRU. (1973) International Commission on Radiation
Units and Measurements. Measurement of Absorbed
Dose in a Phantom Irradiated by a Single Beam of X or
Gamma Rays. ICRU Report 23 (Bethesda, MD: Interna-
tional Commission on Radiation Units and Measure-
ments).
ICRU. (1980) International Commission on Radiation
Units and Measurements. Radiation Quantities and
Units. ICRU Report 33 (Bethesda, MD: International
Commission on Radiation Units and Measurements).
ICRU. (1984) International Commission on Radiation
Units and Measurements. Stopping Powers for Elec-
trons and Positrons. ICRU Report 37 (Bethesda, MD:
International Commission on Radiation Units and
Measurements).
ICRU. (1985a) International Commission on Radiation
Units and Measurements. Dose and Volume Specifica-
tion for Reporting Intracavitary Therapy in Gynecology.
ICRU Report 38 (Bethesda, MD: International Commis-
sion on Radiation Units and Measurements).
ICRU. (1985b) International Commission on Radiation
Units and Measurements. Determination of Dose Equi-
valents Resulting from External Radiation Sources.
ICRU Report 39 (Bethesda, MD: International Commis-
sion on Radiation Units and Measurements).
ICRU. (1986) International Commission on Radiation
Units and Measurements. Modulation Transfer Func-
tion of Screen-FilmSystems. ICRUReport 41 (Bethesda,
MD: International Commission on Radiation Units and
Measurements).
ICRU. (1987) International Commission on Radiation
Units and Measurements. Use of Computers in External
Beam Radiotherapy Procedures with High-Energy
Photons and Electrons. ICRU Report 42 (Bethesda,
MD: International Commission on Radiation Units and
Measurements).
ICRU. (1988) International Commission on Radiation
Units and Measurements. Determination of Dose Equi-
valents fromExternal Radiation SourcesPart 2. ICRU
Report 43 (Bethesda, MD: International Commission on
Radiation Units and Measurements).
ICRU. (1989a) International Commission on Radiation
Units and Measurements. Clinical Neutron
DosimetryPart I: Determination of Absorbed Dose in
a Patient Treated by External Beams of fast Neutrons.
ICRU Report 45 (Bethesda, MD: International Commis-
sion on Radiation Units and Measurements).
ICRU. (1989b) International Commission on Radiation
Units and Measurements. Tissue Substitutes in Radi-
ation Dosimetry and Measurement. ICRU Report 44
(Bethesda, MD: International Commission on Radiation
Units and Measurements).
ICRU. (1992a) International Commission on Radiation
Units and Measurements. Measurements of Dose Equi-
valents from External Photon and Electron Radiations.
ICRU Report 47 (Bethesda, MD: International Commis-
sion on Radiation Units and Measurements).
ICRU. (1992b) International Commission on Radiation
Units and Measurements. Phantoms and Computa-
tional Models in Therapy, Diagnosis and Protection.
ICRU Report 48 (Bethesda, MD: International Commis-
sion on Radiation Units and Measurements).
ICRU. (1992c) International Commission on Radiation
Units and Measurements. Photon, Electron, proton
and Neutron Interaction Data for Body Tissues. ICRU
Report 46 (Bethesda, MD: International Commission on
Radiation Units and Measurements).
ICRU. (1993a) International Commission on Radiation
Units and Measurements. Prescribing, Recording and
Reporting Photon Beam Therapy. ICRU Report 50
(Bethesda, MD: International Commission on Radiation
Units and Measurements).
ICRU. (1993b) International Commission on Radiation
Units and Measurements. Quantities and Units in
Radiation Protection Dosimetry. ICRU Report 51 (Beth-
esda, MD: International Commission on Radiation
Units and Measurements).
ICRU. (1995) International Commission on Radiation
Units and Measurements. Medical ImagingThe
Assessment of Image Quality. ICRU Report 54 (Beth-
esda, MD: International Commission on Radiation
Units and Measurements).
ICRU. (1997) International Commission on Radiation
Units and Measurements. Dose and Volume Specifica-
tion for Reporting Interstitial Therapy. ICRU Report 58
(Bethesda, MD: International Commission on Radiation
Units and Measurements).
ICRU. (1998a) International Commission on Radiation
Units and Measurements. Conversion Coefficients for
Use in Radiological Protection against External Radi-
ation. ICRU Report 57 (Bethesda, MD: International
Commission on Radiation Units and Measurements).
ICRU. (1998b) International Commission on Radiation
Units and Measurements. Clinical Proton
DosimetryPart I: Beam Production, Beam Delivery
and Measurement of Absorbed Dose. ICRU Report 59
(Bethesda, MD: International Commission on Radiation
Units and Measurements).
ICRU. (1998c) International Commission on Radiation
Units and Measurements. Fundamental Quantities
and Units for Ionizing Radiation. ICRU Report 60
REFERENCES
107
(Bethesda, MD: International Commission on Radiation
Units and Measurements).
ICRU. (1999) International Commission on Radiation
Units and Measurements. Prescribing, Recording and
Reporting Photon Beam Therapy (Supplement to ICRU
Report 50). ICRU Report 62 (Bethesda, MD: Interna-
tional Commission on Radiation Units and Measure-
ments).
ICRU. (2001) International Commission on Radiation
Units and Measurements. Dosimetry for High Energy
Photon Beams based on Standards of Absorbed Dose to
Water. ICRU Report 64. J. ICRU 1 (1) (Bethesda, MD:
International Commission on Radiation Units and
Measurements).
ICRU. (2002) International Commission on Radiation
Units and Measurements. Absorbed Dose Specification
in Nuclear Medicine. ICRU Report 67. J. ICRU 2 (1)
(Bethesda, MD: International Commission on Radiation
Units and Measurements).
ICRU. (2003) International Commission on Radiation
Units and Measurements. Image Quality in Chest
Radiography. ICRU Report 70. J. ICRU 3 (2) (Bethesda,
MD: International Commission on Radiation Units and
Measurements).
ICRU. (2004a) International Commission on Radiation
Units and Measurements. Prescribing, Recording, and
Reporting Electron BeamDosimetry. ICRUReport 71. J.
ICRU 4 (1) (Bethesda, MD: International Commission
on Radiation Units and Measurements).
ICRU. (2004b) International Commission on Radiation
Units and Measurements. Dosimetry of Beta Rays and
Low-Energy Photons for Brachytherapy with Sealed
Sources.. ICRU Report 72. J. ICRU 4 (2) (Bethesda,
MD: International Commission on Radiation Units and
Measurements).
IEC. (1996) International Electrotechnical Commission.
Medical Electrical EquipmentDosimetric Instruments
for Non-Invasive Measurements of X-Ray Tube Voltage
in Diagnostic Radiology, IEC Publication 61676
(Geneva: International Electrotechnical Commission).
IEC. (1997) International Electrotechnical Commission.
Medical Electrical EquipmentDosemeters with Ion-
ization Chambers and/or Semiconductor Detectors as
Used in X-Ray Diagnostic Imaging. IEC Publication
61674 (Geneva: International Electrotechnical Commis-
sion).
IEC. (1999) International Electrotechnical Commission.
Medical Electrotechnical Equipment-Particular
Requirements for the Safety of X-Ray Equipment for
Computed Tomography. IEC International Standard
60601-2-44 (Geneva: International Electrotechnical
Commission).
IEC. (2000) International Electrotechnical Commission.
Dose Area Product Meters. IEC Publication 60580
(Geneva: International Electrotechnical Commission).
IEC. (2004) International Electrotechnical Commission.
Medical Diagnostic X-Ray EquipmentRadiation Con-
ditions for Use in the Determination of Characteristics.
IEC Publication 61267, FIDS (Geneva: International
Electrotechnical Commission).
Iles, W. J. (1987) The Computation of Bremsstrahlung X-
ray Spectra over an Energy Range 15 keV to 300 keV.
NRPB-R204 (London, UK: Her Majestys Stationary
Office).
IPEM. (1991) Institute of Physics and engineering in
Medicine. Data for Estimating X-Ray Tube Total Filtra-
tion. IPEMReport no. 64 (York, UK: Institute of Physics
and Engineering in Medicine).
IPEM. (1997) Institute of Physics and engineering in
Medicine. Catalogue of Diagnostic X-Ray Spectra and
Other Data. IPEM Report 78 (York, UK: Institute of
Physics and Engineering in Medicine).
IPSM. (1991) Institute of Physical Sciences in Medicine.
Report of the IPSM working party on low- and medium-
energy x-ray dosimetry. Phys. Med. Biol. 36,
10271038.
IPSM, NRPB and CoR. (1992) Institute of Physical Sci-
ences in Medicine, National Radiological Protection
Board and Royal College of Radiographers. National
Protocol for Patient Dose Measurements in Diagnostic
Radiology. (Chilton, UK: NRPB).
ISH and EC. (1998) Institut fu r Strahlen Hygiene and
European Commission. Reference doses and quality in
medical imaging. Radiat. Prot. Dosim. 80, 1354.
ISO, IEC, OIML and BIPM. (1992) International Organ-
ization for Standardization. International Electrotech-
nical Commission, International Organization of Legal
Metrology and International Bureau of Weights and
Measures. Guide to the Expression of Uncertainty in
Measurement. (Geneva: International Organization for
Standardization).
Jacobson, A. F., Cameron, J. R., Siedband, M. P. and Wag-
ner, J. (1976) Test cassette for measuring peak tube
potential of diagnostic x-ray machines. Med. Phys. 3,
1925.
Jansen, J. T. M., Dierker, J. and Zoetelief, J. (1994)
Calculation of air kerma to mean glandular dose
conversion factors for mammography units employing
various target-filter combinations. In: Proceedings of
the Xth Scientific Symposium of the Belgian Society of
Hospital Physicists, Schaeken, B. and Vanregemorter,
J., Eds. (Antwerpen, Belgium: Belgian Society of Hos-
pital Physicists) pp. 6675.
Jansen, J. T., Geleijns, J. and Zoetelief, J. (1995)
Calculation of computed tomography dose index to
effective dose conversion factors based on measurement
of the dose profile along the fan shaped beam. Br. J.
Radiol. 69, 3341.
Jones, D. G. and Wall, B. F. (1985) Organ Doses from
Medical X-Ray Examinations Calculated Using Monte
Carlo Techniques. NRPB-R186 (London, UK: Her
Majestys Stationary Office).
Jones, D. G. and Shrimpton, P. C. (1991) Survey of CT
Practice in the UK. Part 3: Normalised Organ Doses
Calculated Using Monte Carlo Techniques. NRPB-
R250 (London, UK: Her Majestys Stationary Office).
Jones, D. G. and Shrimpton, P. C. (1993) Normalised
Organ Doses for X-ray Computed Tomography Calcu-
lated Using Monte Carlo Techniques. NRPB-SR250
(Chilton, UK: National Radiological Protection Board).
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
108
Kemerink, G. J., Borstlap, A. C. W., Frantzen, M. J.,
Schultz, F. W., Zoetelief, J. and van Engelshoven, J.
M. A. (2001) Patient and occupational dosimetry in
double contrast barium enema examinations. Br. J.
Radiol. 74, 420428.
Kereiakes, J. G. and Rosenstein, M. (1980) Handbook of
Radiation Doses in Nuclear Medicine and Diagnostic X-
Rays. (Boca Raton, FL: CRC Press, Inc.).
Kerr, G. D. and Eckerman, K. F. (1985) Neutron and
photon fluence-to-dose conversion factors for active
marrow of the skeleton. In: Proceedings of the Fifth
Symposium on Neutron Dosimetry, vol. I, Schraube,
H., Burger, G. and Booz, J., Eds., EUR 9762 EN (Lux-
embourg: CEC). pp. 133145.
Kicken, P. J. H., Kemerink, G. J. and Van Engelshoven,
J. M. A. (1999) Patient dosimetry in arteriography of
the lower limbs. Part I: quantification of exposure.
Radiat. Prot. Dosim. 81, 2536.
Klein, R., Aichinger, H., Dierker, J., Jansen, J. T. M.,
Joite-Barfuss, S., Sa bel, M., Schulz-Wendtland, R.
and Zoetelief, J. (1997) Determination of average
glandular dose with modern mammography units
for two large groups of patients. Phys. Med. Biol.
42, 651671.
Klevenhagen, S. C. (1982) The build-up of backscatter in
the energy range 1 mm Al to 8 mm Al HVT. Phys. Med.
Biol. 27, 10351043.
Klevenhagen, S. C. (1989) Experimentally determined
backscatter factors for x-rays generated at voltages
between 16 and 140 kV. Phys. Med. Biol. 34,
18711882.
Knoll, G. F. (2000) Radiation Detection and Measurement,
3rd edition (New York, NY: John Wiley and Sons, Inc.).
Kramer, H. M., Selbach, H.-J. and Iles, W. J. (1998) The
practical peak voltage of diagnostic x-ray generators.
Br. J. Radiol. 71, 200209.
Kramer, R., Zankl, M., Williams, G. and Drexler G.
(1982) The Calculation of Dose from External Photon
Exposures Using Reference Human Phantoms and
Monte Carlo Methods. Part I: the Male (ADAM) and
Female (EVA) Adult Mathematical Phantoms. GSF Ber-
icht S-885 (Neuherberg: GSF-National Research Centre
for Environment and Health, D-85764, ISSN 0721-
1694).
Kramer, R., Drexler, G, Petoussi-Henss, N., Zankl, M.,
Regulla, D. F. and Panzer, W. (2001) Backscatter fac-
tors for mammography calculated with Monte Carlo
methods. Phys. Med. Biol. 46, 771781.
Ku stner, H. (1924) Qualitatives u ber Filterschutz gegen
Verbrennungen in der Rontgendiagnostik. Fortschr. D.
Rontgenstr. 32, 324342.
Larsson, J. P., Persliden, J., Sandborg, M. and Alm Carls-
son, G. (1996) Transmission ionization chambers for
measurements of air collision kerma integrated over
beam area. Factors limiting the accuracy of calibration.
Phys. Med. Biol. 41, 23812398.
Larsson, J. P., Persliden, J. and Alm Carlsson, G. (1998)
Ionization chambers for measuring air kerma integ-
rated over beam area. Deviations in calibration values
using simplified calibration methods. Phys. Med. Biol.
43, 599604.
Lesperance, J. (1982) Coronary AngiographyProjections.
Radiology Department Report (Montreal, Canada:
Institut de Cardiologie de Montreal).
Ma, C and Nahum, A. E. (1995) Calculations of ion cham-
ber displacement effect corrections for medium-energy
x-ray dosimetry. Phys. Med. Biol. 40, 4562.
Ma, C.-M., Coffey, C. W., DeWerd, L. A., Liu, C., Nath, R.
Seltzer, S. M. and Seuntens, J. P. (2001) AAPMprotocol
for 40-300 kVx-ray beamdosimetry in radiotherapy and
radiobiology. Med. Phys. 28, 868893.
Maccia, C., Benedittini, M., Lefaure, C. and Fagnani, F.
(1988) Doses to patients from diagnostic radiology in
France. Health Phys. 54, 397408.
Maccia, C., Wall, B. F., Padovani, R., Shrimpton, P. C. and
Husson B. (1989) Results of a trial set up by a study
group of the radiation protection programme of the
CEC. In: Optimization of Image Quality and Patient
Exposure in Diagnostic Radiology, BIR Report 20,
Moores B. M., Wall B. F., Eriskat, H. and Schibilla, H.,
Eds. (London: British Institute of Radiology).
Marshall, N. W., Noble, J. and Faulkner, K. (1995) Patient
and staff dosimetry in neuroradiological procedures.
Br. J. Radiol. 68, 495501.
McDowell, R. B., Morgan, T. J. and Smith, L. (1976)
Experimental Procedure for the Evaluation of the X-
Ray Exposure Study Model. (Rockville, MD: Bureau of
Radiological Health).
McKeever, S. W. S., Moscovitch, M. and Townsend, P. D.
(1995) Thermoluminescence Dosimetry Materials: Prop-
erties and Uses. (Ashford, Kent, UK: Nuclear Techno-
logy Publishing).
McKinlay, A. F. (1981) Thermoluminescent Dosimetry. In:
Medical Physics Handbook, vol. 5 (Bristol, UK: Adam
Hilger).
McMaster, W. H., Kerr, G. D., Del Grande, B., Mallet, J. H.
and Hubbell, J. H. (1969) Compilation of X-Ray Cross-
Sections. Report UCRL-50174 Sec 2, Rev 1 (Springfield,
USA: NTIS).
McParland, B. J. (1998) Entrance skin dose estimates
derived from dose-area product measurements in inter-
ventional radiological procedures. Br. J. Radiol. 71,
12881295.
MCPLIB. (1988) Monte Carlo Photon Library. Cross-
Section Data of the MCNP Code Based on Hubbell et al.
(1975) and Everett and Cashwell (1973) (Los Alamos,
NM: Los Alamos National Laboratory).
Miller, D. L., Balter, S., Noonan, P. T. and Georgia, J. D.
(2002) Minimizing radiation-induced skin injury in
interventional radiology procedures. Radiology 225,
329336.
Morgan, R. H. (1961) The measurement of radiant energy
levels in diagnostic roentgenology. Radiology 76,
867877.
NAS/NRC (1980) National Academy of Sciences/National
Research Council. The Effects on Populations of Expos-
ure to Low Levels of lonizing Radiation.. Committee on
the Biological Effects of Ionizing Radiations. BEIR I
(Washington, DC: National Academy Press).
NAS/NRC (1990) National Academy of Sciences/National
Research Council. Health Effects of Exposure in Low
Levels of Ionizing Radiation. NCS Report 10 (Delft,
REFERENCES
109
Netherlands: Netherlands Commission on Radiation
Dosimetry).
NCRP. (1986) National Council on Radiation Protection
and Measurements. MammographyA Users Guide.
NCRP Report No. 85 (Bethesda, MD: National Council
on Radiation Protection and Measurements).
NCS. (1997) Dosimetry of Low- and Medium Energy X
Rays. NCS Report 10 (Delft, Netherlands: Netherlands
Commission on Radiation Dosimetry).
NEXT. (1978) Nationwide Evaluation of X-ray Trends. US
Dose Surveys 19551975. HEW Publication FDA (Rock-
ville, MD: Bureau of Radiological Health).
Ng, K.-H., Rassiah, P., Wang, H.-B., Hambali, A. S.,
Muthuvellu, P. and Lee, H.-P. (1998) Doses to patients
in routine X-ray examinations in Malaysia. Br. J.
Radiol. 71, 654660.
Oberhofer, M. and Scharmann, A. (1981) Thermolumines-
cent Dosimetry. (Published for the CEC, Bristol, UK:
Adam Hilger).
ODea, T. J., Geise, R. A. and Ritenour, E. R. (1999) The
potential for radiation-induced skin damage in inter-
ventional neuroradiological procedures: a review of
522 cases using automated dosimetry. Med. Phys. 26,
20272033.
Pani, R. Ballesio, P. L., Pellegrini, R. and Soluri, A. (1987)
Organ doses in mathematical phantoms with different
elemental composition of organs and tissues. Nucl.
Instrum. Meth. Phys. Res. A255, 360364.
Panzer, W., Scheurer, C. and Zankl, M. (1989) Dose
to patients in CT examinations: results and con-
sequences from a field study in the Federal Republic of
Germany. In: Optimization of Image Quality and
Patient Exposure in Diagnostic Radiology. BIR Report
20, Moores, B. M., Wall, B. F, Eriskat, H. and Schibilla,
H., Eds. (London, UK: British Institute of Radiology)
pp. 185188.
Papageorgiou, E., Vardalaki, E., Hourdakis, C. J. and
Dimitriou, P. (2001) Estimation of doses received by
patients undergoing radiological examinations in
Greece. Radiat. Prot. Dosim. 93, 3141.
Pernicka, F., Alm Carlsson, G., Dance, D. R., DeWerd, L.
A., Kramer, H. M. and Ng, K.-H (2001) Development of
an International Code of Practice for Dosimetry in X-
Ray Diagnostic Radiology. In: Radiological Protection
of Patients in Diagnostic and Interventional Radiology,
Nuclear Medicine and Radiotherapy, IAEA-CSP-7/P
(Vienna: IAEA) pp. 9398
Peterson, L. and Rosenstein, M. (1989) Computer Program
for Tissue Doses in Diagnostic Radiology (for VAX and
IBM-Compatible PC Systems) (Rockville, MD: Center
for Devices and Radiological Health).
Petoussi-Henss, N., Zankl, M., Drexler, G., Panzer, W. and
Regulla, D. (1998) Calculation of backscatter factors for
diagnostic radiology using Monte Carlo methods. Phys.
Med. Biol. 43, 22372250.
Petoussi-Henss, N., Zankl, M., Fill, U. and Regulla, D.
(2002) The GSF family of voxel phantoms. Phys. Med.
Biol. 47, 89106.
Poll, V., Beddoe, A. H., Aspden, P. J. and King, S. D. (1972)
Thermoluminescence dosimetry of beta emitters in
bone. In: Proceedings of the Second International Con-
ference on Strontium Metabolism. USAEC CONF-
720818 (Springfield, VA: National Technical Informa-
tion Service).
Preston, D. L., Shimizu, Y., Pierce, D. A., Suyama, A. and
Mabuchi, K. (2003) Studies of mortality of atomic bomb
survivors. Report 13: Solid cancer and noncancer dis-
ease mortality: 19501997. Radiat. Res. 160, 381407.
Regulla, D. F., Hieber, L. B. and Seidenbusch, M. (1998)
Physical and biological interface dose effects in tissue
due to x-ray-induces release of secondary radiation from
metallic gold surfaces. Radiat. Res. 150, 92100.
Robertson, J., Holubinka, M. R., Jones, A. P., Moretti, C.
J., Rawlings, D. J., Roberts, P. J. and Wall, B.F (1992)
Calibration of Dosemeters used in Diagnostic Radiology.
A protocol produced by the Dosimetry Working Party of
the Diagnostic Radiology Topic Group of the Institute of
Physical Sciences in Medicine. Scope 1, No. 4.
Rosenstein, M. (1976) Organ Doses in Diagnostic Radi-
ology. HEW Publication FDA 76-8030 (Rockville, MD:
Bureau of Radiological Health).
Rosenstein, M. (1984) Handbook of glandular tissue doses
in mammography. Presentation at the 29th Meeting of
the Health Physics Society, New Orleans, LA. (Rock-
ville, MD: Center for Devices and Radiological Health).
Rosenstein, M. (1988) Handbook of Selected Tissue Doses
for Projections Common in Diagnostic Radiology. HHS
Publication FDA 89-8031 (Rockville, MD: Center for
Devices and Radiological Health).
Rosenstein, M., Warner, G. G. and Beck, T. J. (1979)
Handbook of Selected Organ Doses for Projections Com-
mon in Pediatric Radiology. HEW Publication FDA 79-
8079 (Rockville, MD: Bureau of Radiological Health).
Rosenstein, M., Andersen, L. W. and Warner, G. G. (1985)
Handbook of Glandular Tissue Doses in Mammo-
graphy. HHS Publication FDA 85-8239 (Rockville, MD:
Center for Devices and Radiological Health).
Rosenstein, M., Suleiman, O. H., Burkhart, R. L.,
Stern, S. H. and Williams, G. (1992) Handbook of
Selected Tissue Doses for the Upper Gastrointestinal
Fluoroscopic Examination. HHS Publication FDA 92-
8282 (Rockville, MD: Center for Devices and Radiolo-
gical Health).
Sa ez, J. C., Romero, A. M., Fera ndez, C., Gomez, S.,
Va zquez, J. and Olivares, M. P. (1999) Dose measure-
ments in a phantom simulating neonates by using dif-
ferent TL materials: LiF: Mg, Cu, P and LiF: Mg, Ti.
Radiat. Prot. Dosim. 85, 345348.
Sankaranarayanan, K and Chakraborty, R. (2000)
Ionizing radiation and genetic risks. XIII. Summary
and synthesis of papers VI to XII and estimates of gen-
etic risks in the year 2000. Mutat. Res. 453, 183197.
Schultz, F. W., Geleijns, J. and Zoetelief, J. (1994)
Calculation of dose conversion factors for posterior-
anterior chest radiography of adults with a relatively
high x-ray spectrum. Br. J. Radiol. 67, 775785.
Schultz, F. W., Jansen, J. T. and Zoetelief, J. (2001) Organ
and Effective Dose Conversion Factors for for Various
Types of Examination in Diagnostic Radiology. State-
chreport: ST-MSF-2001-012 (Delft, The Netherlands:
Interfaculty Reactor Institute Delft University of Tech-
nology).
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
110
Seelentag, W. W., Panzer, W., Drexler, G., Platz, L. and
Santer, F. (1979) A Catalogue of Spectra for the Calib-
ration of Dosemeters. GSF-Bericht S 560 (Neuherberg:
GSF-National Research Centre for Environment and
Health, D-85764, ISSN 0721-1694).
Seuntjens, J and Verhaegen, F. (1996) Dependence of the
overall correction factor for a cylindrical ionization
chamber on field size and depth in medium-energy
x-ray beams. Med. Phys. 23, 17891796.
Shope, T. B., Gagne, R. M. and Johnson, G. C. (1981) A
method for describing the doses delivered by transmis-
sion x-ray computed tomography. Med. Phys. 8,
488495.
Shrimpton, P. C., Wall, B. F. and Fisher, E. S. (1981) The
tissue-equivalence of the Alderson Rando anthropo-
morphic phantom for x-rays of diagnostic qualities.
Phys. Med. Biol. 26, 133139.
Shrimpton, P.C., Wall, B.F. (1982) An evaluation of the
Diamentor transmission ionization chamber in indicat-
ing exposure-area product (R cm
2
) during diagnostic
radiological examinations. Phys. Med. Biol. 27,
871878.
Shrimpton, P. C., Jones, D. G., Hillier, M. C., Wall, B.
F., LeHeron, J. C. and Faukner, K. (1991) Survey of
CT Practice in the UK Part 2: Dosimetric Aspects.
NRPB-R249 (London, UK: Her Majestys Stationary
Office).
Shrimpton, P.C., Wall, B.F., Jones, D.G., Fisher, E.S.,
Hillier, M.C. and Kendall, G.M. (1986) A National
Survey of Doses to Patients Undergoing a Selection
of Routine X-Ray Examinations in English Hospitals.
NRPB-R200 (Her Majestys Stationary Office, Lon-
don, UK).
Snyder, W. S., Ford, M. R., Warner, G. G. and Fisher, H. L.
Jr (1969) Estimates of absorbed fraction for monoener-
getic photon sources uniformly distributed in various
organs of a heterogeneous phantom. MIRD Pamphlet
No. 5. J. Nucl. Med. 10, (Suppl. 3).
Sovik, E., Klow, N. E., Hellesnes, J. and Lykke, J. (1996)
Radiation-induced skin injury after percutaneous
transluminal coronary angioplastyCase report. Acta
Radiol. 37, 305306.
Spiers, F. W. (1963) Interim report on the determination
of dose to bone marrow from radiological procedures.
Br. J. Radiol. 36, 238240.
Stanton, L., Bratelli, S. D. and Day, L. J. (1982)
Measurements of diagnostic x-ray backscatter by a
novel ion chamber method. Med. Phys. 9, 121130.
Stanton, L., Villafana, T., Day, J. L. and Lightfoot, D. A.
(1984) Dosage evaluation in mammography. Radiology
150, 577584.
Stern, S. H., Rosenstein, M., Renaud, L. and Zankl, M.
(1995a) Handbook of Selected Tissue Doses for Fluoro-
scopic and Cineangiographic Examination of the Coron-
ary Arteries (in SI units). HHS Publication FDA95-8289
(Rockville, MD: Center for Devices and Radiological
Health).
Stern, S. H., Dennis, M. J., Williams, G. and Rosenstein,
M. (1995b) Simulation of the upper gastrointestinal
fluoroscopic examination for calculation of absorbed
dose in tissue. Health Phys. 69, 391395.
Storm, E. and Israel, H. I. (1970) Photon cross sections
from 1 keV to 100 MeV for elements Z1 to Z100.
Nuclear Data Tables Sec A 7, 565688.
Struelens, L. (2005) Optimization of patient doses, linked
to image quality, in vascular radiology, Ph.D. thesis,
Vrije Universiteit Brussel, Brussels, Belgium.
Suleiman, O. H. (1989) Development of a method to calcu-
late organ doses for the upper gastrointestinal fluoro-
scopic examination, Ph.D. Dissertation, School of
Hygiene and Public Health, Johns Hopkins University,
Baltimore, MD.
Tapiovaara, M., Lakkisto, M. and Servomaa, A. (1997)
PCXMC: A PC-Based Monte Carlo Program for Calcu-
lating Patient Doses in Medical X-Ray Examinations.
Report STUK-A139 (Helsinki: Finnish Centre for Radi-
ation and Nuclear Safety).
Taylor, L. S. (1958) 80 Years of quantities and units. Per-
sonal reminescences, Part II: From the International
X-Ray Unit Committee to the International Commis-
sion of Radiation Units and Measurements. ICRU
News 901, 49.
Taylor, L. S. (1990) History of the International Commis-
sion of Radiation Units and Measurements (ICRU).
Health Phys. 1, 306314.
Tipton, I. H., Snyder, W. S. and Cook, M. J. (1966) Ele-
mental Composition of Standard Man. ORNL-4007
(Oak Ridge, TN: Oak Ridge National Laboratory).
Trout, E. D., Kelley, J. P. and Lucas, A. C. (1960)
Determination of half-value layer. Am. J. Rontgenol.
84, 729734.
Tubiana, M., Aurengo, A., Averbeck, D., Bonnin, A., Le
Guen, B., Masse, R., Monier, R., Valleron, A. J. and de
Vathaire, F. (2005) La relation dose-effet et lestimation
des effets canceroge`nes des faibles doses de rayonne-
ments ionisants. Rapport a` lAcademie des Sciences et
a` lAcademie de Medecine, Edit. (Paris: Nucleon).
UNSCEAR. (1972) United Nations Scientific Committee
on the Effects of Atomic Radiation. Ionizing Radiation:
Levels and Effects, vol. 1 Levels (New York: United
Nations).
UNSCEAR. (1988) United Nations Scientific Committee
on the Effects of Atomic Radiation. Sources Effects and
Risks of Ionizing Radiation (New York: United
Nations).
UNSCEAR. (2000) United Nations Scientific Committee
on the Effects of Atomic Radiation. Sources and
Effects of Ionizing Radiation, Report to the General
Assembly with Scientific Annexes (New York: United
Nations).
UNSCEAR. (2001) United Nations Scientific Committee
on the Effects of Atomic Radiation. Hereditary Effects
of Radiation. UNSCEAR 2001 Report to the General
Assembly with Scientific Annex. (New York: United
Nations).
Van Dam, J. and Marinello, G. (1994) Radiothermolumin-
escent dosimeters. In: Methods for in vivo Dosimetry in
External Radiotherapy. European Society for Thera-
peutic Radiology and Oncology. Booklet No. 1 in the
series Physics for Clinical Radiotherapy, 1st edition
(Leuven: Garant) pp. 3754.
REFERENCES
111
Van de Putte, S., Verhaegen, F., Taeymans, Y. and
Thierens, H. (2000) Correlation of patient skin doses
in cardiac interventional radiology with dose-area
product. Br. J. Radiol. 73, 504513.
Van o, E., Guibelalde, E., Ferna ndez, J. M., Gonza lez, L.
and Ten, J. I. (1997) Patient dosimetry in interven-
tional radiology using slow films. Br. J. Radiol. 70,
195200.
Van o, E., Arranz, L., Sastre, J. M., Moro, C., Ledo, A.,
Garate, L. T. and Minguez, I. (1998) Dosimetric and
radiation protection considerations based on some
cases of patient skin injuries in interventional cardi-
ology. Br. J. Radiol. 71, 510516.
Van o, E., Goicolea, J., Galvan, C., Gonza lez, L., Meiggs, L.,
Ten, J. I. and Macaya, C. (2001a) Skin radiation injur-
ies in patients following repeated coronary angioplasty
procedures. Br. J. Radiol. 74, 10231031.
Van o, E., Gonzalez, L., Ten, J. I., Fernandez, J. M., Guibe-
lalde, E. and Macaya, C. (2001b) Skin dose and dose-
area product values for interventional cardiology
procedures. Br. J. Radiol. 74, 4855.
Van o, E., Ferna ndez, J. M., Ten, J. I., Guibelalde, E.,
Gonza lez, L. and Pedroza, C. S. A. (2002) Real-time
measurement and audit of radiation dose to patients
undergoing computed radiography. Radiology 225,
283288.
Van Sonnenberg, E., Shankar, S., Morrison, P. R., Nair, R.
T., Silverman, S. G., Jaklitsch, M. T., Liu, F., Cheung,
L., Tuncali, K. Skarin, A. T. and Sugarbaker, D. J.
(2005) Radiofrequency ablation of thoracic lesions:
Part 2, initial clinical experienceTechnical and mul-
tidisciplinary considerations in 30 patients. Am. J.
Roentgenol. 184, 381390.
Veit, R., Zankl, M., Petoussi, N., Mannweiler, E., Williams,
G. and Drexler, G. (1989) Tomographic Anthropo-
morphic Models. Part 1: Construction Technique and
Description of Models of an 8 Week Old Baby and a 7
Year Old Child. GSF-Bericht 3/89 (Neuherberg: GSF-
National Research Centre for Environment and Health,
D-85764, ISSN 0721-1694).
Veit, R. and Zankl, M. (1992) Influence of patient size on
organ doses in diagnostic radiology. Radiat. Prot.
Dosim. 43, 241243.
Veit, R. and Zankl, M. (1993) Variation in organ doses in
pediatric radiology due to patient diameter, calculated
with phantoms of varying voxel size. Radiat. Prot.
Dosim. 49, 353356.
Vij, D. R., Ed. (1993) Thermoluminescent Materials
(Englewood Cliffs, NJ: PTR Prentice Hall).
Villard, M. P. (1908) Instruments de mesure a lecture
directe. Arch. dElec. Med. 16, 692.
Wagner, L. K. and Pollock, J. J. (1999) Real-time monit-
oring to estimate dose to skin of patients from high dose
fluoroscopy. Br. J. Radiol. 72, 846855.
Waite, J. C. and Fitzgerald, M. (2001) An assessment of
methods for monitoring entrance surface dose in fluoro-
scopically guided interventional procedures. Radiat.
Prot. Dosim. 94, 8992.
Wall, B. F. (1989) Quality control of dose-area product
meters. In: Technical and Physical Parameters for
Quality Assurance in Medical Diagnostic Radiology,
BIR Report 18, Moores, B. M., Stieve, F. E., Eriskat H.
and Schibilla H., Eds. (London, UK: British Institute of
Radiology) pp. 140142.
Wambersie, A., Smeesters, P. and Fruhling, J. (1995)
Exposition aux rayonnements ionisants: effets radio-
biologiques et pathogenie. Louvain Medical
114, S97S132.
Wambersie, A., Menzel, H. G., Gahbauer, R. A., Jones, D.
T. L., Michael, B. D. and Paretzke, H. G. (2002)
Biological weighting of absorbed doses in radiation
therapy. Radiat. Prot. Dosim. 99, 445452.
Wambersie, A., Zoetelief, J., Menzel, H. G. andParetzke, H.
(2006) The ICRU (International Commission on
Radiation Units and Measurements): its contribution
to dosimetry in diagnostic and interventional radiology.
Radiat. Prot. Dosim. Advance Access published January
31, 2006.
Warner, G. G. (1973) BRHGAM: A Medical X-Ray Dose
Estimation Program. ORNL-TM-4393 (Oak Ridge, TN:
Oak Ridge National Laboratory).
Wise, K. N., Sandborg, M., Persliden, J. and Alm Carlsson,
G. (1999) Sensitivity coefficients for converting
entrance surface dose and air-kerma area product to
effective dose and energy imparted to the patient.
Phys. Med. Biol. 44, 19371954.
Wochos, J. F., Detorie, N. A. and Cameron, J. R. (1979)
Patient exposure from diagnostic x-rays: an analysis of
1972-1975 NEXT data. Health Phys. 36, 127134.
Wu, D.-K., Sun, F.-Y. and Dai, H.-C. (1984) A high sensit-
ivity LiF thermoluminescent dosemeter-LiF (Mg, Cu,
P). Health Phys. 46, 22372250.
Wu, X. (1991a) Breast dosimetry in screen-film mammo-
graphy. In: Screen-Film Mammography: Imaging Con-
siderations and Medical Physics Responsibilities,
Barnes, G. T. and Tucker, D. M., Eds. (Madison, WI:
Medical Physics Publishing) pp. 159175.
Wu, X., Barnes, G. T. and Tucker, D. M. (1991b) Spectral
dependence of glandular tissue dose in screen-film
mammography. Radiology 179, 143148.
Wu, X., Gingold, E. L., Barnes, G. T. and Tucker, D. M.
(1994) Normalized average glandular dose in
molybdenum target-rhodium filter and rhodium
target-rhodium filter mammography. Radiology 193,
8389.
Yamada, M., Wong, F. L., Fujiwara, S., Akahoshi, M. and
Suzuki, G. (2004) Noncancer disease incidence inatomic
bomb survivors, 19581998. Radiat. Res. 161, 622632.
Young, K. C., Ramsdale, M. L. and Bignall, F. (1998)
Review of dosimetric methods for mammography in
the UK breast screening programme. Radiat. Prot.
Dosim. 80, 183186.
Zankl, M. and Wittmann, A. (2001) The adult male voxel
model Golem segmented from whole body CT patient
data. Radiat. Environ. Biophys. 40, 153162.
Zankl, M., Veit, R., Williams, G., Schneider, K., Fendel, H.,
Petoussi, N. and Drexler, G. (1988) The construction of
computer tomographic phantoms and their application
in radiology and radiation protection. Radiat. Environ.
Biophys. 27, 153-164.
Zankl, M., Petoussi, N., Veit, R. and Fendel, H. (1989)
Organ doses for a child in diagnostic radiology:
PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING
112
comparison of a realistic and a MIRD-type phantom.
In: Optimization of Image Quality and Patient
Exposure in Diagnostic Radiology, BIR Report 20,
Moores, B. M., Wall, B. F., Eriskat, H. and
Schibilla, H., Eds. (London: British Institute of Radi-
ology,) pp. 196198.
Zankl, M., Panzer, W. and Drexler, G. (1991) The Calcu-
lation of Dose from External Photon Exposures Using
Reference Human Phantoms and Monte Carlo Methods.
Part VI: Organ Doses from Computed Tomographic
Examinations. GSF-Bericht 30/91 (Neuherberg: GSF-
National Research Centre for Environment and Health,
D-85764, ISSN 0721-1694).
Zankl, M., Petoussi, N. andDrexler, G. (1992) Effective dose
equivalentthe impact of the new ICRP definition
for external photon irradiation. Health Phys. 62,
395399.
Zankl, M., Panzer, W. and Drexler, G. (1993) Tomographic
Anthropomorphic Models. Part II: Organ Doses from
Computed Tomographic Examinations in Paediatric
Radiology. GSF-Bericht 30/93 (Neuherberg: GSF-
National Research Centre for Environment and Health,
D-85764, ISSN 0721-1694).
Zoetelief, J. and Jansen, J. T. (1995) Calculation of
air kerma to average glandular tissue dose conversion
factors for mammography. Radiat. Prot. Dosim. 57,
397400.
Zoetelief, J., Julius, H. W. and Christensen, P.
(2000) Recommendations for Patient Dosimetry in
Diagnostic Radiology Using TLDs (Luxembourg:
Office for Official Publications of the European Commu-
nities).
Zoetelief, J., Julius, H. W. and Christensen, P. (2003a)
Recommendations for patient dosimetry in diagnostic
radiology using thermoluminescence dosimetry. In:
Standards and Codes of Practice in Medical Radiation
Dosimetry. Proceedings of an International Symposium,
Vienna, 2528 November 2002, IAEA-CN-96/39.
(Vienna: International Atomic Energy Agency).
pp. 439447.
Zoetelief, J., Pernicka, F., Alm Carlsson, G., Dance, D. R.,
DeWerd, L. A., Drexler, G., Ja rvinen, H., Kramer, H. M.
and Ng, K.-H. (2003b) Dosimetry in Diagnostic and
Interventional Radiology: International Commission
on Radiation Units and Measurements and IAEA activ-
ities. In: Standards and Codes of Practice in Medical
Radiation Dosimetry, Proceedings of an International
Symposium, Vienna, 2528 November 2002, IAEA-CN-
96/39. (Vienna: International Atomic Energy Agency)
pp. 387404.
REFERENCES
113

Você também pode gostar