A Wireless Sensor Network Based Health Monitoring Infrastructure and Testbed
The health-care system in developed countries will increasingly come under pressure as the average age of their
We envision Ayushman to meet the following set of goals:
With Ayushman we are developing a dependable and secure health monitoring infrastructure that can collect, query and analyze patient health information in real-time. Ayushman's conceptual architecture organizes its essential features as a set of interacting services. There are essential 5 components to this organizations:
Ayushman's Conceptual Architecture
In the previous section we presented the conceptual architecture of Ayushman. In our current implementation however, Ayushman is organized as a three-tiered hierarchical cluster. At the lowest tier are the medical devices and sensors which collect data from the patients. The data collected by them is sent to their cluster head node called the local gateway implemented on a PDA. The PDAs form the second-tier cluster and are lead by an external gateway node implemented on a laptop and these laptops form the third-tier headed by a workstation designated the central server which manages the entire system . Each of the tier has been designed to be self contained and can survive on its own even in the event of disconnection and network partition.
Localization Capabilities Integrated into Ayushman
As a part of the Ayushman project implementation we also integrated a localization system, to provide location based medical services and access control on the systems information. The localization system was initially built to provide location based access control to computing resources in our lab. It basically uses a scene analysis based approach to localize an entity. We implemented and tested this system within our lab, as a first step we trained system by creating a signal strength based map of our lab and then used scene analysis to calculate various locations within our lab. Then when an actual entity is detected, based on the scene analysis results and signal strength of its id, it can be localized.
Implementation of the Initial Prototype
EKG and Blod Pressure Monitoring: We implemented Ayushman, using off-the-shelf components and integrated them together. Firstly we integrated an EKG monitor (Vernier EKG) and Blood-Pressure Monitor (Suntech Advantage BP monitor). These two devices do not provide any wireless communication capability, therefore they were connected to a Mica2/MicaZ mote using a serial port. The motes controlled them and provided a wireless interface to transmit their data which is eventually stored in a central server.
Oximeter: The Oximeter Sensor and board (Smith Medical) provides us with the ability to monitor in real-time a person blood oxygen level, photoplethysmogram and heart rate. The device does not provide any wireless communication capability, therefore was interfaced with a MicaZ mote using a serial port. We implemented a simple PCB board allowed us to connect the device directly to the mote through the 51pin header. The PCB performed the necessary switching between the logic levels of RS232 and MicaZ mote.The design was obtained from the CodeBlue projects. The motes controlled them and provided a wireless interface to transmit their data which is eventually stored in a central server.
Gait Monitoring: We have implemented a gait monitoring system as a part of the Ayushman system. Gait monitoring is done using accelrometer, gyroscopic sensors attached to MicaZ motes. Further we also use piezoelectric heal and toe strike sensors to detect each step. The gait analysis is done at the PC based on the data sent by these sensors.
Environment Monitoring: Further we also collect ambient information from around the patient, including temperature, humidity, along with light information. This information is corellated with the EKG, BP and Gait Monitoring sensors to obtain an accurate state of the patient's health. Environmental monitoring is done using TelosB sensors which also provided a wireless interface to transmit their data to a central server.
Software Components: Most of the backend software for Ayushman has been written in C#. The local gateway is implemented on three platforms - Dell Axim PDA and Motorola Cellphone which collects physiological data from the sensors, stores it and forwards to the central server. We have also experimented with a Crossbow Stargate platform as the local gateway with equally good results. The central sever stores all the data it receives on a MS Access database and allows remote clients to connect, view and query patient data. Priliminary experiments on the lifetime of the motes and and the PDA show that they can last for upto a month and 14 hours respectively before batteries have to be recharged.
Undergraduate Student Projects
Last Updated: 11th May 2008
Photos: AFP, companiontechnologies.com,saddsask.ca
Sponsor logos are registered trademarks of National Science Foundation and Mediserve Information Systems Inc, respectively