This thesis addresses the hypothesis that unobtrusive monitoring of daily living could be used to evaluate changes in the health status of frail elderly people living alone at home. Low cost motion sensors can be used for monitoring the long term trends in occupant?s well-being in terms of physical, mental, social and environmental factors. The monitored data can be used to quantitatively measure parameters that can provide insight into the level of activity and functional ability of the subject. Any deviations in these parameters can provide information on the changing health status of the subject. This thesis attempts for the first time to establish a mathematical and statistical framework for the monitoring of functional health status in the home using a network of wireless sensors to monitor occupancy in each room of the house. A low power and low cost, unobtrusive occupancy monitoring system using ZigBee wireless technology and passive infrared sensors has been developed by the Biomedical Systems Laboratory at the University of New South Wales. The essence of the occupancy monitoring system is to detect variations in the activities of daily living (ADL) of elderly people living alone at home. The finite state, discrete parameter, time homogeneous Markov chain represents a theoretical framework for an unobtrusive occupancy monitoring system. An implementation of this framework for monitoring occupancy pattern is presented in real time use. The system was evaluated in a series of field studies in laboratory and home environment, in supervised and unsupervised settings, using cohorts of healthy elderly subjects living alone in their homes in community dwelling setting. A pilot trial was conducted in which four healthy elderly subjects living alone had PIR motion sensors installed in their homes at strategic points for a period of up to 13 weeks. The functionality of the system was evaluated over a domain of basic daily activities. A profile of the activities, in real time environment, for different times and days was stored as transition probability matrices. Automatic techniques for interpreting the test data captured by the system in terms of human movements were evaluated and compared with the wellness profile of the subject. Trial results exhibited that clinically significant model parameters were able to detect longitudinal deviations in the functional health status of elderly people.
| Identifer | oai:union.ndltd.org:ADTP/240719 |
| Date | January 2008 |
| Creators | Kaushik, Alka Rani, Electrical Engineering & Telecommunications, Faculty of Engineering, UNSW |
| Publisher | Publisher:University of New South Wales. Electrical Engineering & Telecommunications |
| Source Sets | Australiasian Digital Theses Program |
| Language | English |
| Detected Language | English |
| Rights | http://unsworks.unsw.edu.au/copyright, http://unsworks.unsw.edu.au/copyright |
Page generated in 0.0016 seconds