IDENTIFICATION OF SIMULATED ANKLE SPRIAN FROM COMMON SPORTS MOTIONS BY FFT AND POLYNOMIAL FITTING ON DORSAL FOOT KINEMATICS DATA

Abstract

This study presented a method to identify simulated ankle sprain motion from common sporting motions by using motion sensor data. Five subjects performed 100 simulated sprain trials and 245 non-sprain common sporting motion trials in a laboratory. Eight motion sensors were attached to different anatomical landmarks on the right foot to collect 3-D linear acceleration and angular velocity data, which was then processed by fast Fourier transform and polynomial fitting in frequency domain. The coefficients of fitted curve were compared, and the accuracy to identify ankle sprain motion achieved 97.0%. Comparing to a previous method, the current method requires shorter data collection and computation time, and could serve for real time ankle sprain monitoring.