Neural Network for On-bed Movement Pattern Recognition

• Published in: 2018 11th Biomedical Engineering International Conference (BMEiCON) pp. 1 - 4
• Main Authors: Sri-ngernyuang, Chawakorn, Youngkong, Prakarnkiat, Lasuka, Duangruedee, Thamrongaphichartkul, Kitti, Pingmuang, Watcharapong
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01-11-2018
• Subjects: Biological neural networks; Hospitals; Neural Networks and Fall Prevention; On-Bed Pattern; Pattern recognition; Tactile sensors
• DOI: 10.1109/BMEiCON.2018.8609998

One of the critical issues in hospitals is the injury from falling out of patient bed. Some of these cases lead to death. Considering this type of incident, a monitoring and alarming system called NEFs (Never Ever Falls) is introduced to prevent patients from falling out of the bed. In this paper, on-bed pattern recognition is developed by applying Neural Network Pattern Recognition from MATLAB. In the experiment, data from 6 persons in 5 different on-bed patterns (Sitting inside the bed, Supine, Lateral on the left, Lateral on the right and sitting at bedsides and corners) is recorded. According to the confusion matrix, training and validation confusion tables show 99.5% and 89.1% accuracy, respectively.