Detection of Fear of Falls using PPG with Video Stimulation of the Fall Condition

Abstract

One of the problem of the after stroke walking training is fear of fall. Fear of falling is significantly corresponded to earlier falls and poor physical function. Lower high frequency (HF) power which is measured from heart rate variability has correlation to the stress, panic, anxiety, or worry. This study proposed a method to detect fear of falling using Photoplethysmography (PPG). The PPG signal is acquired using Arduino. The acquired signal is then processed using several step including trend removal, band frequency filtering, band power estimation, and decision system. Trend removal is conducted using moving average while the band frequency filtering is conducted using second order Butterworth band pass filter. The 2 section video is used to induce relax and fear of falling on the subject feeling which is used as the class on decision system. The output of the decision system is then compared to the class and the MAPE is computed. The result shows that the system can yield MAPE 16.63% which is can be categorized as a good result.