Optimizing Gait Parameters and Insole Sensor Positioning for Parkinson's Disease Assessment

Abstract

Gait abnormality is a characteristic symptom of Parkinson's disease (PD) and could be exploited to assess PD progression. In this study, we examined various gait parameters and insole sensor positioning for evaluating PD. We first verified the results from several published papers in which gait parameters exhibited significant differences between PD patients and healthy controls. Then, we investigated additional gait parameters derived from individual sensors in 8 positions across the sole. The result demonstrated that the balls, heels, and center of arches are valuable positions for PD gait assessment. Furthermore, a random forests method showed the most important gait parameters to predict PD include swing time on the balls and medial arches, double support time on the balls and medial arches, and ground reaction forces on the heels. The optimization of sensor positioning and gait parameters suggests a low-cost and effective way identify Parkinsonian gait characteristics.