Escolar Documentos
Profissional Documentos
Cultura Documentos
Introduction
is right himself, but also make the patient cooperate. In this paper, we describe a
novel smart mote-based portable medical system which automatically monitors and
handles medication among persons with dementia at homes based on wireless
multimodal sensors and mobile phone or PDA technology to promote medication
adherence. It provides assistive cues to patients in the form of prompts and reminders,
and allows physicians or caregivers to monitor patients medicine taking activity and
obtain summarized behavioral reports from their PDA anytime, anywhere. Section 2
discusses the related works. Section 3 describes the design considerations and details
of a smart medical system. Section 4 describes the PDA or mobile phone applications
to the caregivers and doctors for caregiving and clinical assessment in a context aware
fashion. Finally, section 5 concludes with a discussion of future works.
Related Works
Previously, many systems have been developed to support medication adherence. The
assistive technology lab in University of Toronto proposes a medication reminding
system [1] using context aware technology while ETH Zurich proposes a smart
medicine cabinet [2] using passive RFID and Bluetooth-enabled active tags to
monitor the contents of the box. The Lanchaster University [3] designs a device to
support the management of medication in a community care environment, and the
University of Ulster proposes a pill container [4] for medication. Recent works in
Intel by Jay Lundell [5] proposes a smart context aware medication adherence system.
Our work is similar to them in that we also use sensors, context aware systems and
PDA or mobile phones to provide assistive cue to patients such as reminders for
medical compliance. However, our work is different in some ways as we seek to
provide a single holistic integrated portable smart medical system to collectively
address and satisfy the different needs and perspectives of all stakeholders for
medication compliance such as the patients, caregivers and doctors using the
pervasive mote platform. In the long term, we hope to integrate more sophisticated
behavior understanding system to provide holistic solutions to dementia patients
beyond medication adherence using the popular mote platform.
In this section, we will describe our hardware and software design considerations of a
smart medical system for monitoring and handling medication in dementia persons.
3.1
Design Considerations
We study the requirements from the perspectives of patients, caregivers and doctors,
and also from the literature survey. In all cases, it requires the smart medical system
to be safe but reliable enough to capture the medication taking behavior of patient.
Good recognition rate must be achieved and false alarms should be minimized to
System Design
The smart medical system is designed to meet the above considerations and consists
of a medication box with sensors, patient medication analyzer, central server and
PDA or mobile phones. It is designed such that a smart medicine box can be
connected to multiple related doctors and caregivers, and vice versa.
The medicine box periodically sends sensor readings wirelessly to the patient
medication analyzer. Using a PDA, a doctor can first authorize himself through the
central server, and connect to the patient medication analyzer to obtain behavioral
reports of the patient in a distributed manner. This system is illustrated in Figure 1.
The smart medical system provides scalable monitoring provisioning and support
standardized schemes for automated intervention management and activity planning.
The detailed hardware and software components are described in the next sections.
3.1.2
Hardware
The wireless smart mote-based medicine box for medical adherence among persons
with dementia is designed as shown in Figure 2.
Fig. 2. The smart medicine box implemented with various sensor technologies
It utilizes multimodal sensors and actuators to monitor and assist patient in taking
medicine. The smart medicine box consists of nine drawers seven drawers for each
day in a week which allows the patient to differentiate the medicine to be taken for a
particular day and two general ones. A LED will blink to assist the patient in opening
the right drawer. If a wrong compartment is opened, an error tone will be generated to
alert the patient. In case the patient forgets to take the medicine, a reminder tone will
be generated to remind the patient it is time to take his medicine. The details of the
hardware components are described below:
Motion Sensing Medicine Box Lid and Environmental Light Sensing
A mote with accelerometer and light sensor is attached to the main lid of the medicine
box to sense whether it is opened or not. The lid is detected as open if the
accelerometer sensor readings drop by an amount greater than 25 and if the light
sensor detects a light intensity greater than 700 (maximum reading is 1024).
Sensing Human Presence and Motion (PIR)
An external PIR sensor connected to the mote is used to sense the presence of the
patient around the box. Output of the PIR sensor is connected to the ADC port on the
sensor board. When motion is detected, this output readings will go below 600. An
ultrasonic sound sensor is also used to determine the distance of patient from the
medicine box. Signal from an external ultrasonic sensor powered by a 6V voltage
source is fed into the ADC1 port of the MTS101CA sensor board on the mote. The
ADC port interprets the ultrasonic sensor readings in a range from 0 to 1024. If main
lid is closed, the distance to obstacle will be very small. When the lid is opened, the
value will be very high (>200). Using a simple algorithm, the system reports the
motion of patient in front when the readings are in the range from 20 to 100.
Controlling LED Indicators
When the main lid is opened, LED on the drawer to be opened will start blinking.
Since there are only limited motes and PWR output ports available, we design a
circuit so that the 7 LEDs can be turned on or off efficiently. 3 output ports on the
sensor board of this mote, PW3, PW4, PW5, are used to control the LED indicators as
shown in Figure 3. The 3 output voltages from the mote are connected to a 3 to 8
decoder. The outputs from the decoder are connected to the negative input of the 7
LEDs. The positive terminals of every LEDs are also connected to the output pin of a
555 timer, which generates a high and low voltage alternatively.
Tone Generation
Tones are generated by two output voltages to the buzzer/speaker when a mote
receives a control message from the patient medication analyzer. The controlling
circuit is depicted in Figure 5 below.
3.1.3
Software
Experiments are conducted by students to simulate patients behavior and the results
are encouraging that we will try to deploy the system in patients home for evaluation.
While development is still in its early stages, our joint effort with a local hospital
should see us achieving our objective of validating it in a real life setting.
5 Conclusion
We present a smart medical system for monitoring and handling medicine taking
behaviors for dementia patients, caregivers and doctors. The use of multimodal
sensors, actuators and PDA or mobile phone is the first step for us to promote medical
compliance, and we are now furthering our work by including more sensor modalities
such as pressure sensors, RFID, etcr. to enhance the recognition rate and reduce false
alarm, and also working on adaptive interface such as LED status board for patients.
The joint effort with a local hospital should see us achieving our long term objective
of integrating more sophisticated behavior understanding system to provide holistic
solutions to dementia patients beyond medication adherence using the mote platform.
References
1. Mihailidis, et al., A context-aware medication reminding system: Preliminary design and
development. Rehabilitation Engineering and Assistive Technology Society of North
America, Atlanta, Georgia, CD-ROM Proceedings
2. Matthias Lampe, et al., Advances in Pervasive Computing, Austrian Computer Society
(OCG). Vienna, Austria, April (2004)
3. Kember, et al., Designing Assistive Technologies for Medication Regimes in Care Settings,
Universal Access in the Information Society (2003);2:235-242
4. Nugent, et al., Can Technology Improve Compliance to Medication, 3rd International
Conference on Smart Homes and Health Telematics (2005)
5. L. Jay et al., Why Elders Forget to Take Their Meds: A Probe Study to Inform a Smart
Reminding System, 4Th International Conference on Smart Homes and Health Telematics,
(2006)