IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 49, NO.
6, JUNE 2002
Movement Quantification in Epileptic Seizures:
A New Approach to Video-EEG Analysis
Zhanjian Li*, António Martins da Silva, and João Paulo Silva Cunha, Member, IEEE
Abstract—It is common that epileptic seizures induce uncoor-
dinated movement in a patient’s body. This movement is a rele-
vant clinical factor in seizure identification. Nevertheless, quan-
tification of this information has not been an object of much at-
tention from the scientific community. In this paper, we present
our effort in developing a new approach to the quantification of
movement patterns in patients during epileptic seizures. We attach
markers at landmark points of a patient’s body and use a camera
and a commercial video-electroencephalogram (EEG) system to
synchronously register EEG and video during seizures. Then, we
apply image-processing techniques to analyze the video frames and
extract the trajectories of those points that represent the course of
the quantified movement of different body parts. This information
may help clinicians in seizure classification. We describe the frame-
work of our system and a method of analyzing video in order to
achieve the proposed goal. Our experimental results show that our
method can reflect quantified motion patterns of epileptic seizures,
which cannot be accessed by means of traditional visual inspection
of video recordings. We were able, for the first time, to quantify
the movement of different parts of a convulsive human body in the
course of an epileptic seizure. This result represents an enhanced
value to clinicians in studying seizures for reaching a diagnosis.
Index Terms—Epilepsy, movement quantification, video-EEG.
I. INTRODUCTION
I T IS common that epileptic seizures induce uncoordinated
movement in a patient’s body. This movement is a relevant
clinical factor in seizure identification. Nevertheless, quantifi-
cation of this information has not been an object of much at-
tention from the scientific community. Traditionally, long-term
video-electroencephalogram (EEG) monitoring is employed to
study these cases [1], [2]. Although movement is very important
in the diagnosis of epilepsy, for many years, researchers have
concentrated on developing quantification algorithms to extract
EEG features which are invisible to traditional visual inspection.
Similar approaches have not been carried out on video informa-
tion, where clinicians only perform visual inspection. This vi-
sual analysis allows only a rough estimate of the motion pattern
Manuscript received July 10, 2001; revised December 12, 2001. This work
was supported by the Portuguese Science and Technology Foundation (F.C.T.)
under PRAXIS XXI and POSI programs, grants BD/13603/97 and Sapiens
33802/99. Asterisk indicates corresponding author.
*Z. Li is with the Department of Electronics and Telecommunication/IEETA,
University Campus of Santiago, University of Aveiro, 3810-193 Aveiro, Por-
tugal (e-mail: li@ieeta.pt).
A. Martins da Silva is with the Neurophysiology Department, S. António
General Hospital, 4099-001 Porto, Portugal. He is also with ICBAS, Univer-
sity of Porto, 4099-001 Porto, Portugal.
J. P. Silva Cunha is with the Department of Electronics and Telecommu-
nication/IEETA, University of Aveiro, 3810-193 Aveiro, Portugal. He is also
with the Neurophysiology Department, S. António General Hospital, 4099-001
Porto, Portugal (e-mail: jcunha@det.ua.pt).
Publisher Item Identifier S 0018-9294(02)04846-2.
0018-9294/02$17.00 © 2002 IEEE
565
of the seizure to be made. To our knowledge, no effort has been
made to develop quantification algorithms that extract informa-
tion from video recorded seizures. The main reasons for this are,
on the one hand, the traditional video monitoring units rely on
analogs video recording devices, which are an obstacle to the
use of image analysis approaches and, on the other hand, cap-
turing, storing, and processing video data demands enormous
throughput from computers. Recently, the emergence of full dig-
ital video-EEG technology based on the support of computer
multimedia has opened up such possibilities.
We developed a new approach to quantify patient movement
during epileptic seizures based on a commercial video-EEG
system. Our motivation was propted by the following two facts.
• Clinicians consider uncoordinated seizure motion pattern
(or motor semiology) of the patient’s body as valuable ev-
idence in the identification of the presence of a seizure
and of its source [1], [3]. Traditional visual inspection of
long-term video-EEG recordings cannot evaluate quanti-
fied motion information.
• Computer technology is in constant development. Dig-
ital video systems and video-EEG equipment are being
applied progressively more in the clinical diagnosis of
epilepsy.
These facts have enabled us to promote a line of research that
aims to study movement patterns of epileptic seizures, by means
of computer technology. From the clinical point of view, this
research addresses the following questions.
• Can quantified movement analysis extract information
that is imperceptible to the visual inspection of an epilep-
tologist?
• Can movement quantification information be employed
to provide additional evidence in the classification of
epileptic seizures?
• What kind of quantified movement information is helpful
in the diagnosis of epilepsy?
• Do these techniques contribute toward a more reliable di-
agnosis?
In this paper, we propose a new experimental approach to start
addressing these questions.
II. RELATED WORK
Extracting human motion has been an important research
topic in computer vision for many years. However, analyzing
the motion of the human body is a challenging problem because
the human body is a highly flexible structure that can twist
and bend in many degrees of freedom. In addition, the variety
of luminance, noise induced by devices and self-occlusions of
the human body complicate the movement analysis process.
566
To address this complex problem, some investigators used
marker-based methods to evaluate human body motion for
clinical diagnosis. For instance, Jobbágy et al. [4] used a
marker-based method to analyze movement patterns of a pa-
tient hand for aiding in early detection of Parkinson’s disease.
These authors attached markers to the middle phalanges of a
patient’s hand and extracted marker trajectories for evaluation.
In their experiments, the subject was requested to put his hands
on the desk and pinch, twiddle, and tap. A camera captured
this procedure and the vertical displacements of markers were
extracted and evaluated. Miyazaki et al. [5] developed a system
to quantify the eyelid movement for clinical purposes. They
attached markers to the lower margin of the upper eyelid of
a subject. The subject was requested to blink and the marker
movements were recorded and evaluated. Motion quantification
can also be performed without markers. Nakajima et al. [6]
used an image optical flow analysis approach to quantitatively
evaluate body motion and visualize the apparent velocity field
of the entire body motion, including both chest respiratory
movement and posture change.
Although these approaches are suitable for their target appli-
cations, they cannot be used to extract quantified information
for more complex human motion. For instance, in the work car-
ried out by Jobbágy et al. [4], the subject’s hands are required
to be placed steadily on the desk so that the relating positions of
the markers (fingers) do not change, and the markers are not oc-
cluded. This constraint is very important because the highlight
markers are easily identified according to their horizontal po-
sitions to enable the establishment of motion correspondence.
However, this approach will fail in complex movements such as
rotation, crossing, etc. The approaches that are based in optical
flow, of which the work of Nakajima et al. [6] is an example,
only work in the case of simple motion due to their high sensi-
tivity to noise.
Some researchers used model-based approaches for tracking
or recognizing human motion such as walking [7], [8] or other
simple motions [9], [10]. However, these approaches can only
work for a few simple human motions. It is difficult to use
them in our target application because the movement induced
by epileptic seizures is complex and uncoordinated and is, thus,
very difficult to model and quantify. From our literature review,
marker-based methods seemed to be the most suitable for our
targeted application.
In order to attain our primary goal, we attach some markers
to the landmark points of a patient’s body so that the motion
of the markers reflects the movement of underlying body parts.
A method is proposed to obtain quantified information of com-
plex human motion. We focus our attention on knowing which
parts of the human body move and what are their dynamics. This
strategy simplifies the procedure of extracting motion patterns
by avoiding the need to address topics such as gesture recogni-
tion or geometry restructure.
III. METHODS
A. The Framework
In order to ensure clinical acceptance, we designed our
approach according to the routine video-EEG procedure used
in clinical neurophysiology environments, which is similar in
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 49, NO. 6, JUNE 2002
every center. Usually, a patient lies on a bed and is monitored
by a video camera and an EEG machine. Our experimental
setup is illustrated in Fig. 1.
A monochrome CCD camera with several infrared light emit-
ting diodes parallel to the lens are mounted right above the bed
at an approximate distance of 1.5 m. A patient lies on the bed.
The image size is adjusted to bring the whole body of the sub-
ject into focus. A set of infrared reflective markers is attached to
landmark points of the body. This is done through black elastic
braces where these markers are attached in order to enhance the
contrast in the image. The light sources in the ceiling are in-
candescence lamps which are kept at relatively low intensity in
order that highlighted blobs are formed in the video image only
by infrared reflective markers.
We assume that the motion brought on by the seizure cannot
cause the patient to fall out of view, such as falling off the bed.
Hence, the camera is kept stationary without tilting or rotating
during the video capture. A commercial video-EEG system [11]
is used to synchronously acquire video sequences and EEG data
into a PC. The video data is saved in a video file that can then be
analyzed by means of a software tool we have developed, called
quantification of movement in epileptic seizures (QMovES).
B. Marker Definition
Due to the originality of our approach and to the consequent
lack of established standards, we defined a marker positioning
system with 22 landmark points (MPS22) on the human body
as illustrated in Fig. 2.
The markers should be attached to the landmark points be-
fore the video acquisition. They are covered with infrared re-
flective material so that they reflect this light invisible to human
vision. They are very small and lightweight not blocking the
patient movement. An infrared-sensitive CCD camera shows
the markers as high intensity clusters of pixels (blobs) in video
frames that are detectable through our method.
C. Extraction of Quantified Information of Motion
QMovES imports the video file as input of its quantified
movement extracting method. This method is divided in two
stages: initialization and tracking, as illustrated in Fig. 3.
1) Initialization: This stage includes calculating a
threshold, segmenting, and assigning labels to blobs as
well as calculating initial parameters used in the tracking stage.
Due to the framework we are using, the intensity of pixels
of highlighted blobs is higher than pixels of other objects in
the scene. To segment them, we calculate a grey level threshold
from a frame based on an approach purposed by Johannsen [12]
using an entropy measure. This threshold is used to segment the
blobs in video frames in the tracking stage, assuming that the
indoor luminance does not change too much during video-EEG
acquisition. After segmentation, we calculate the geometric cen-
troid of the blobs. In this step, we assume that clustered pixels
in a small search window belong to one blob. This constraint
is intuitively similar to human perception. A scan is performed
to find all the blobs in the image. Given our application do-
main, we cannot assume what the patient’s initial posture was
and even cannot assume that all defined markers are exposed
to view. Thus, it is very difficult to establish a complete marker
LI et al.: MOVEMENT QUANTIFICATION IN EPILEPTIC SEIZURES 567

Fig. 1. The setup of QMovES system. A CCD camera acquires video sequences into a PC where a digital video-EEG system and the QMovES software tool are
installed.

In our method we model the movement of a marker as a two-

dimensional (2-D) ballistic motion [13]. A marker position in
two successive frames can be approximately expressed by
(1)
(2)
where and denote the centroid coordinate of the marker,
and denote instantaneous velocity along the axis direc-
tions, and the corresponding instantaneous accelerations,
is frame number, and denotes a small temporal sampling
interval that is equal to 1/(frame rate). Assuming we have the
three adjacent frames , , and , then we can estimate
and and and as follows:
(3)

(4)

Fig. 2. The 22 landmark positions we defined for the full-body marker (5)
positioning system (MPS22).

positioning system necessary for automatic labeling, because no
prior knowledge about the patient’s posture is available for ini-
tialization and training purposes. Given these facts, we perform
the initial labeling procedure interactively by means of a mouse
click sequence.
(6)
Given this model, we use the first three frames of seizure
recordings to calculate the initial parameters according to
(3)–(6) which are necessary to initialize a Kalman filter [14]
for prediction purposes in the tracking phase.
568
Fig. 3. Schematic representation of the procedure for movement quantification
used in QMovES.
2) Tracking: The tracking stage of our method uses the
information obtained from the initialization stage, and iterates
it for every remaining frame in the sequence to compute the
successive positions of the markers. In this procedure, a video
frame is first extracted and segmented as described previously.
Then the clusters of highlight pixels (blobs) are extracted using
the algorithm proposed by Jobbágy [15]. Following this, marker
positions are predicted and used in the motion correspondence
estimation section of the method. This is a key part of our
method. To provide the reader with a better understanding of
this problem and of the solution we have designed for it, we
present an illustration of a situation where this problem occurs
in Fig. 4. We assume the points (labeled with A, B, C, D, and E)
move at different velocity and acceleration. The opened circles
denote the points in one frame, which are already labeled, and
the filled circles denote the points in the next frame. We overlap
two frames in one figure to show how the position changes. The
problem at this point is how we label the filled circle points in
Fig. 4(a) and is technically named “correspondence problem”
[16].
As referred previously, we model the movement by (1) and (2)
and use a Kalman filter to predict marker positions in the next
video frame. These predicted positions, presented in Fig. 4 as
squares, are then used in motion correspondence estimation by
a distance criteria optimization algorithm based on a normalized
distance that we define as
(7)
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 49, NO. 6, JUNE 2002
(a) (b)
Fig. 4. Illustrations of the “correspondence problem” where the point’s
motion from one frame to next are represented. The opened circles denote
the points in the first frame and the filled circles denote the points in the
next frame. The squares are predicted positions for the points. (a) Overlap of
two frames to show the positions change. (b) Denotes actual correspondence
solution computed (arrows).
where is the number of markers, the number of blobs
found in the current frame, and is the pre-
dicted position of th marker; is
the position of th blob found, denotes the magnitude of
vector , and denotes the th frame. The blobs are identified
as correspondent to respective markers by minimizing the sum
of the normalized distance between the predictive and the ef-
fective positions of the blobs. This normalized distance already
takes into consideration speed, direction and acceleration be-
cause marker positions are predicted by (1) and (2). The de-
nominator denotes the sums of absolute quantities for all pos-
sible matches. Hence, the proposed measure gives the relative
measure of distance from predicted position to actual position.
Large displacements will lead to large values.
Although this proposed solution proved to work correctly, the
correspondence problem is combinatorial explosive. If there are
points in a sequence of frames, resulting in trajectories,
the number of possible trajectory sets is which may
lead to a highly complex and computational intensive solution.
To cope with this complexity we implemented in QMovES a
modified version of a noniterative Greedy algorithm originally
proposed by Rangarjan et al. [16] where (7) is used as cost func-
tion. This algorithm deals with video frame by frame in a loop,
in which the outputs are the positions of markers in the cur-
rent frame. These outputs are imported into the Kalman filters
to iteratively predict parameters for the next loop. Through this
procedure, the correspondence is solved as shown in Fig. 4(b).
In our application domain the set of positions of the markers
(trajectories) reflect the quantified motion information of the
underlying body parts. The movement patterns can then be ex-
tracted from this information.
IV. RESULTS
Fig. 5(a) is an original video frame taken under our frame-
work in which the highlight blobs on the frame denote the
markers attached to the patient’s body. Fig. 5(b) shows the
segmentation result where the blobs are extracted from the
background. Fig. 5(c) shows the centroid positions of blobs
(cross markers) found by the algorithm described in the
LI et al.: MOVEMENT QUANTIFICATION IN EPILEPTIC SEIZURES
(a)
(c)
Fig. 5. (a) An original video frame with highlight blobs, (b) the segmentation result, (c) detection and correspondence of the blobs, (d) the output of tracking
+
stage with predicted positions ( ) and actual positions ( ).
previous section. Fig. 5(d) shows an output of a tracking pro-
cedure. The squares denote the predicted positions of markers
generated by the Kalman filter. The crosses denote the actual
positions of markers located by the motion correspondence
estimation part of our method.
The output trajectories of those markers from this video se-
quence are showed in Fig. 6. Fig. 6(a) shows the motion of the
markers (pixel unit) in axis direction and Fig. 6(b) shows sim-
ilar information in the axis direction. In Fig. 6(c), the speed
of the markers is depicted. If we know the mass of the under-
lying body parts, this data can be converted to show the effective
quantity of movement involved in each section of the convulsive
body.
We decided to display the movement-quantified information
by mimicking EEG tracings in order to ease the adaptation of
the clinician to the reading this information and to facilitate the
correlation between EEG and movement events in an easy way.
The marker trajectories are then the quantified motion informa-
tion, which denotes the motion pattern of different body parts.
For instance, from Fig. 6(c), we can learn that the legs (K1, K2,
J1, J2, I1, and I2) move intensively from second 6 to second 21.
In Fig. 6(b), from the tracings of A1, A2, B1, and B2, we can
measure the behavior of the head that shows six rotations from
second 5 to second 21. One important achievement is that this
information is given to the user in a quantified way. We are able
to register movement events and quantify their speed and spatial
motion patterns. This type of quantification may make a lot of
difference in clinical application and is apparently achieved for
the first time.
To enhance the pattern recognition capability, we also visu-
alize the motion pattern by an animated color-coded graphic.
569
(b)
(d)
This type of visualization technique is commonly used in neu-
rophysiology to reconstruct the underlying potentials of EEG or
their specific frequency band’s power distribution in the scalp.
Once more, it was our intention to ease the clinicians’ adapta-
tion to the reading of this new type of information by using a
familiar approach to present it.
In Fig. 7 we show several frames of an animated color-coded
graphic of quantified movement processed by a median filter.
The highlighted areas represent the intensity of movement of
those body parts. The scale in the left bar denotes the motion
speed in pixels/s. From Fig. 7, we can learn the motion dynamic
pattern during the seizure and characterize its evolution. For in-
stance, we can quantify the hand’s motion at time 00:06:280
and the leg’s intense motion at times 00:07:120, 00:07:360, and
00:10:000. We can also analyze the intensity of the head move-
ment observed at times 00:10:000 and 00:20:760. This infor-
mation is made available as a video sequence to dynamically
visualize the motion pattern exhibited by the patient during a
seizure.
V. DISCUSSION
Given the originality of the approach we have presented, we
are not able to compare it with other studies in the literature. We
can identify strong advantages in its application but also have
to face several problems that should occupy our efforts in the
course of this line of research in the near future.
The digital video-EEG technique is a high input–output
throughput application which is characterized by large data
sets that usually are only bearable by computer systems when
applying high compression to the original data sets. However,
570
X
Fig. 6. (a) Tracings of the markers motion in coordinate in function of time with the corresponding body positions. (b) Tracings of the markers motion in
coordinate in function of time with the corresponding body positions.
a high compression rate leads to a loss of image information
and degradation in the quality of the image. In practice, to save
disk space, the compression rate should be kept high while
video quality goes to a relatively low (but acceptable) level.
In our study, the video data is saved in M-JPEG format with a
compression rate of 16 : 1. Our approach proved to be highly
robust to compression noise and can work correctly with rela-
tively low-quality images. In our system setup, the intensity of
the markers is almost uniform for different indoor luminance.
Under the condition of low luminance, the intensity of a diffuse
background is always lower than that of the markers even if
the patient is wearing white clothes. In this case, the algorithm
can work properly. However, when luminance is increased, the
luminosity of the background will also increase, which may
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 49, NO. 6, JUNE 2002
(a)
(b)
Y
cause difficulties in image segmentation. For robust image
segmentation we used dark colored background objects and
low-medium luminance. We also recommended that patients
wear dark colored clothes if possible.
A difficulty in the tracking procedure of our approach arises
when some markers may become hidden from view due to oc-
clusion by other body parts during the movement. This case is
more serious if the patient lies sideways or in a prone position on
the bed. In these situations there is no prediction model available
and the marker detection algorithm cannot work properly. Nev-
ertheless, in cases of temporary occlusion, we can visually esti-
mate positions and assign labels for them manually. This work
can be done simply by mouse click. The benefit of this proce-
dure is that the occluded points in the frame will have approxi-
LI et al.: MOVEMENT QUANTIFICATION IN EPILEPTIC SEIZURES 571

(c)
Fig. 6. (Continued.) (c) Tracings of the markers’ speed in function of time with the corresponding body positions.

Fig. 7. Frames from an animated color-coded video showing the dynamics of the quantity of movement in the course of a seizure. This procedure shows very
impressive results in the perception of movement patterns.

mately correct positions that still contribute to the trajectories in
the final output. However, this method is not valid for the cases
of long-term occlusion.
The markers may be lost during the tracking phase if more
than two markers move too close, although this situation
is relatively rare when compared with occlusion. QMovES
may also label markers wrongly if a large displacement or an
abrupt change in velocity is registered. These cases violate the
correspondence assumptions so that the optimal match does not
correspond to the correct one. Some heuristic approaches may
also be used to respond to these problems. For instance, the four
markers attached to the face of the subject will never change
their relative positions. As those approaches have proven to
be quite effective, we have inserted some heuristic rules in
our method. If the heuristic automated decision does not solve
these problems, the process may be paused (automatically
or manually) and the user may identify the blobs by simple
mouse click. Automatic labeling can then continue after this
interactive interruption.
One limitation of QMovES is the lack of deepness infor-
mation. However, our experiments showed that our quantified
movement information extracted from 2-D images is still able
572
to reflect the motion pattern of the human body. One other in-
convenience is that we must attach markers to the patient, which
is an additional task to standard clinical video-EEG acquisition
procedure. To simplify the procedure of attaching markers, we
fixed the markers to elastic braces. These elastic braces can be
attached to body parts in a few minutes. In mild seizures the
markers usually do not change their relative positions according
to the MPS22 model of Fig. 2. Nevertheless, in some more vi-
olent ones, this may happen. We are already considering other
solutions to avoid this problem.
VI. CONCLUSION AND FUTURE WORK
This work has investigated an original approach to extract
quantified motion information in the course of epileptic
seizures. We attached markers to patients’ bodies and per-
formed the video-EEG procedure in a standard clinical
neurophysiology environment. A method to extract the trajec-
tories of markers is also presented. Our approach has proved to
be able to extract quantified motion trajectories of human body
parts that reflect the motion pattern of patients during epileptic
seizures. These trajectories can be reviewed as multichannel
tracings which are similar to EEG. Through further processing,
we can visualize the motion pattern in the form of an animated
color-coded graphic. These ways of visualizing seizure motion
are also original and proved to be very effective in the clinical
trials we have performed. Our experimental results show
that the quantified motion information intuitively reflects the
movement patterns of epileptic seizures. These quantified
patterns cannot be accurately estimated through traditional
visual inspection of video recordings. We intend to search
for more accurate quantification methods, namely to employ
multicamera video acquisition to extract three-dimensional
information from which we may obtain deepness and also
reduce the cases of occlusion even in cases where a patient
lies sideways or in a prone position. We were able, apparently
for the first time, to observe the course of the motions of
an epileptic seizure in a quantified way. This constitutes an
important result of our work. We are now able to register
movements of different parts of the convulsive human body
and quantify their speed, quantity of movement and 2-D
spatial motion patterns. This type of quantification may make
a lot of difference in clinical epileptology. We hope this new
approach will aid clinicians to find more valuable evidences
that may lead to a more reliable seizure classification based on
quantified motor semiology. One immediate application can
be the help in dissociating epileptic from simulated seizures
(also called pseudoseizures). The possibility to correlate this
quantified motion information with the EEG events occurring
in the course of the seizure or during small movements opens
a completely new research item that we intend to explore in
the near future. Finally, this technique has a lot of potential
to study other human pathological states, namely movement
disorders or behavioral changes which are accompanied by
motor components.
For an interactive overview of our work, readers can access
our web site at http://www.ieeta.pt/sias.
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 49, NO. 6, JUNE 2002
ACKNOWLEDGMENT
The authors acknowledge the two awards this work has al-
ready received from the European Epilepsy Academy and the
Portuguese League Against Epilepsy. They would also like to
thank Dr. T. Roberto for the revision of the manuscript.
REFERENCES
[1] L. F. Quesney and P. Gloor, “Localization of epileptic foci,” in
Long-Term Monitoring in Epilepsy (EEG Suppl. no. 37), J. Gotman, J.
R. Ives, and P. Gloor, Eds., 1985.
[2] M. C. Salinsky, “A practical analysis of computer based seizure de-
tection during continuous video-EEG monitoring,” Electroencephalogr.
Clin. Neurophysiol., vol. 103, pp. 445–449, 1997.
[3] H. Lüders, J. Acharya, C. Baumgartner, S. Benbadis, A. Bleasel,
R. Burgess, D. S. Dinner, A. Ebner, N. Foldvary, E. Geller, H.
Hamer, H. Holthausen, P. Kotagal, H. Morris, H. J. Meencke, S.
Noachtar, F. Rosenow, A. Sakamoto, B. J. Steinhoff, I. Tuxhorn, and
E. Wyllie, “Semiological seizure classification,” Epilepsia, vol. 39, pp.
1006–1013, 1998.
[4] Á. Jobbágy, E. H. Furnée, P. Harcos, M. Tárczy, I. Krekule, and L.
Komjáthi, “Analysis of movement patterns aids the early detection of
Parkinson’s disease,” in Proc. 19th IEEE/EMBS, Chicago, IL, Oct.
30–Nov. 2 1997, pp. 1760–1763.
[5] S. Miyazaki, A. Ishida, and A. Komatsuzaki, “A clinically oriented
video-based system for quantification of eyelid movement,” IEEE
Trans. Biomed. Eng., vol. 47, pp. 1088–1096, Aug. 2000.
[6] K. Nakajima, A. Osa, T. Maekawa, and H. Miike, “Evaluation of body
motion by optical flow analysis,” Jpn. J. Appl. Phys., pt. 1, vol. 36, no.
5A, pp. 2929–2937, 1997.
[7] D. Hogg, “Model-based vision: A program to see a walking person,”
Image Vis. Computing, vol. 1, no. 1, pp. 5–20, 1983.
[8] H. J. Lee and Z. Chen, “Knowledge-guided visual perception of 3-D
human gait from a single image sequence,” IEEE Trans. Syst., Man, Cy-
bern., vol. 22, pp. 336–342, Mar.-Apr. 1992.
[9] K. Akita, “Image sequence analysis of real word human motion,” Pattern
Recogn., vol. 17, no. 1, pp. 73–83, 1984.
[10] C. R. Wren, A. Azarbayejani, T. Darrell, and A. P. Pentland, “Pfinder:
Real-time tracking of the human body,” IEEE Trans. Pattern Anal. Ma-
chine Intell., vol. 19, pp. 780–785, July 1997.
[11] ONYX Digital EEG User’s Manual. Heemstede, The Netherlands:
SBI BV, 1998.
[12] G. Johannsen and J. Bille, “A threshold selection method using informa-
tion measures,” in Proc. 6th Int. Conf. Pattern Recognition, 1982, vol.
1, pp. 140–142.
[13] M. Shah, K. Rangarajan, and P. S. Tsai, “Motion trajectories,” IEEE
Trans. Syst., Man, Cybern., vol. 23, pp. 1138–1150, July-Aug. 1993.
[14] A. M. Tekalp, Digital Video Processing. Englewood Cliffs, NJ: Pren-
tice Hall, 1995.
[15] A. Jobbágy, “Fault diagnosis in biomedical engineering application,” ,
Dept. Meas. Inform. Syst., Tech. Univ. Budapest, Budapest, Hungary,
Tech. Rep., 1997.
[16] K. Rangarjan and M. Shah, “Establishing motion correspondence,”
CVGIP: Image Understanding, vol. 54, no. 1, pp. 56–73, 1991.
[17] I. K. Sethi and R. Jain, “Finding trajectories of feature points in a monoc-
ular image sequence,” IEEE Trans. Pattern Anal. Machine Intell., vol.
PAMI-9, pp. 56–73, Jan. 1987.
[18] R. F. Rashid, “Toward a system for the interpretation of moving light
displays,” IEEE Trans. Pattern Anal. Machine Intell., vol. PAMI-2, pp.
574–581, June 1980.
Zhanjian Li received the B.S. degree in electronic
engineering from Shanghai Jiaotong University,
Shanghai, China, in 1987; the M.S. degree in
computer science from Shenyang Institute of Com-
puting Technology (SICT), the Chinese Academy
of Sciences (CAS), Shenyang, China, in 1993;
the Ph.D. degree in electronics engineering and
Telecommunications from the University of Aveiro,
Portugal, in 2002.
He is currently a research staff at the Institute of
Electronics Engineering and Telematics of Aveiro
(IEETA), Aveiro, Portugal. His research interests include image analysis,
object-oriented programming, and multimedia applications.
Dr. Li received the president award from CAS in 1993.
LI et al.: MOVEMENT QUANTIFICATION IN EPILEPTIC SEIZURES
António Martins da Silva received the medical de-
gree and the Ph.D. degree on neurophysiology, from
University of Porto, Portugal.
He is Professor of Human Physiology with the
Medical School at the Biomedical Sciences Institute
“Abel Salazar,” University of Porto. He is the
Director of Clinical Neurophysiology Department,
Santo António Hospital, Porto, and Director of the
Research Unit on Neuro and Psychophysiology
IBMC, University of Porto. His current research
fields/clinical work are mainly on epilepsy, sleep,
and brain function.
573
Joao Paulo Silva Cunha (S’87–M’90) was born
in Porto, Portugal. He received an Electronics
and Telecommunications Engineering Diploma in
1989 and the Ph.D. degree in 1996, both from the
University of Aveiro, Portugal.
In 1996, he joined the Department of Electronics
and Telecommunications at the University of Aviero,
where he is currently an Associate Professor. Previ-
ously, he was a Research Assistant at the Institute
of Biomedical Sciences, University of Porto, Porto,
Portugal. He is an invited Professor in the Masters in
Biomedical Engineering program of the same University. He has been a Vis-
iting Scientist at the University of Florida, Gainsville, FL. and at the “Instituut
voor Epilepsiebestrijding,” Hemestede, The Netherlands. His research has been
devoted to the study of computer-based systems to support diagnosis in clinical
neurophysiology. His other current research interests are telemedicine and mul-
timedia healthcare information systems.
Dr. Cunha is member at the CEN/TC251 normalization body, a member of
the IEEE EMBS, vice-president of the Portuguese Association for Medical In-
formatics, and a member of the Portuguese Biomedical Engineering Society and
of the Portuguese League Against Epilepsy.