Towards a portable human gait analysis & monitoring system

Abstract

Human Gait analysis is useful in many cases, such as, detecting the underlying cause of an abnormal gait, rehabilitation of subjects suffering from motor related diseases such as Parkinson's disease or Cerebral Palsy, improving the athletic performance of sports person etc. However, gait analysis has seen limited usage, especially in developing countries, because of the high cost involved in setting up a gait laboratory. We present a portable gait analysis system using Inertial Measurement Unit (IMU) sensors to collect movement data and a Smart-phone to process it. IMU sensors has gained significant popularity in the last few years as viable option for gait analysis because its low cost, small size and ease of use. Using the accelerometer and gyroscope data from 3 EXL-S3 IMU sensors (on thigh, shank and foot), we measure kinematic angles in the sagittal plane and detect Heel Strike (HT) and Toe Off (TO) events using methods based on [11] and [4] respectively. To measure the accuracy of our system, we compare it with an Optical Gait Analysis system, which is the current gold standard for gait analysis1. We measure the gait parameters for 3 healthy individuals belonging to different age group and achieve an RMSE of 4.739° ± 1.961°, 3.7° ± 3.02° and 4.12° ± 1.21° for Knee Flexion Extension, Ankle Dorsi Flexion respectively and Hip Flexion Extension respectively. We measure the Heel Strike and Toe Off using shank and foot mounted sensor independently. 34.5 ± 28.3 ms and 27.5 ± 32.8 ms is the RMSE for HT calculated by shank and foot sensor w.r.t. optical system respectively. The RMSE for Toe Off is 36.2 ± 36.8 ms and 37.5 ± 35.9 ms for shank and foot sensor w.r.t. optical system respectively.