P1‐429: Adding Intelligence to a Floor‐Based Array Personnel Detector
Harry W. Tyrer, Fadi Muheidat · Alzheimer s & Dementia · 2016
Common reasons for hospitalization of people with Alzheimer’s disease (dementia) are syncope (fainting), fall and trauma (26%) [[1]]. People with dementia have limited ability to maintain and use wearable devices and the privacy concerns of video based system, a passive personnel monitoring system may be a solution. There is value in passively monitoring people in their daily activities. We enhanced the prototype smart carpet [[2]], which is a floor based personnel detector system, to detect falls using computational intelligence components running on a faster but low cost processor. Our hardware front end reads 128 sensors, with sensors output a voltage due to a person walking or falling on the carpet. The processor is Jetson TK1 [[3]], which provides more computing power than before [[2]]. We generated a dataset using volunteers who walked and fell to test our algorithms. Data frames read from the data acquisition system were analyzed using different algorithms. We used different algorithms. We varied the windows size of number of frames, (WS>1) and threshold (TH) to detect falls. We then generated a dataset by applying a set of fall detection algorithms, and used the video recorded for the fall experiments to train a set of classifiers using multiple test options available in the Weka framework [[4]]. We measured the sensitivity and specificity of the system and other metrics for intelligent detection of falls. Results showed that Computational Intelligence techniques detect falls with 96.2% accuracy and 80% sensitivity and 97.8% specificity. In addition to fall detection, we developed a database system and web applications to retain these data for years. We can detect falls with high accuracy and display this data in real-time. The database of all activities in the carpet allow for extensive data analysis any time in the future.