Estimation of Gait Subphase Time Parameters Based on a Human Electrostatic Field Detection System

Abstract

The parameters of gait phases obtained by detecting gait events can provide a robust basis for the pathological diagnosis and rehabilitation of patients with lower limb disorders. In this paper, a method is proposed to divide the inner-stance phase and inner-swing phase by detecting the gait events of heel-off (HO) and swing to the highest position (SHP) in the stepping electrostatic gait signal (EGS) and to estimate the subphase time parameters. The gait events are detected by using the local extreme points of the integrated signal of the EGS. Four subphase temporal parameters are obtained after combining the local extrema of the EGS representing the initial contact (IC) and toe-off (TO) for the division of the inner-stance and inner-swing phases. Compared with the motion capture system (VICON), the detection accuracy of the proposed method reaches 98.5%, and the measurement error is within the range of ±60 ms. Pearson’s coefficient ‘r’ was used to verify the correlation between the four gait subphases estimated by the two methods, and excellent agreement was obtained (r > 0.95, p < 0.05). Most of the measurement results of the four subphase temporal gait parameters are within the 95% confidence interval of the Bland–Altman analysis chart. The repeatability of the proposed method was calculated through intraclass correlation coefficients (ICC) and showed good test-retest reliability (ICC > 0.85, p < 0.01). This method detects gait events and estimates subphase duration with high accuracy and can provide a foundation for the gait analysis of EGSs.