A multi-sensor approach for fall risk prediction and prevention in elderly

Scientific research on smartphone-based fall detection systems has recently been stimulated due to the growing elderly population and their risk of falls. Even though these systems are helpful for fall detection, the best way to reduce the number of falls and their consequences is to predict and prevent them from happening in the first place. To address the issue of fall prevention, in this paper, we propose a fall prediction system by integrating the sensor data of smartphones with a smartshoe. In our previous research, we designed and implemented a pair of sensing shoes (smartshoe) that contained four pressure sensors with a Wi-Fi communication module in each shoe to unobtrusively collect data in any environment. After assimilating the smartshoe and smartphone sensor data, we performed an extensive set of experiments in the lab environment to evaluate normal and abnormal walking patterns. In the smartphone, the system can generate an alert message to warn the user about the high-risk gait patterns and potentially save them from a forthcoming fall. We validated our approach using a decision tree with 10-fold cross validation and found 97.2% accuracy in gait abnormality detection.