Gait Abnormality Detection in Unilateral Trans-Tibial Amputee in Real-Time Gait Using Wearable Setup

The presented study proposes a novel approach to detect gait abnormalities in unilateral trans-tibial (TT) amputees using a wearable setup. The system uses force-sensitive resistors and potentiometers to collect data on the user's gait patterns. A machine-learning algorithm based on Extreme Learning Machines is utilized to classify the gait patterns as normal or abnormal. The system is evaluated on a dataset of healthy and unilateral TT amputees, and the results reveal that the ELM-based classification technique achieved high accuracy, sensitivity, specificity, and F1 score. The proposed wearable gait setup is tested by conducting a standard 6-m walk test, and the collected data is segmented into stance and swing phases. The study also compares various gait parameters of healthy and amputated subjects, and the results show significant asymmetry in the amputated subjects. The proposed setup also detects asymmetry in force distribution under each foot. The study's findings reveal that the proposed wearable gait setup is a reliable and effective tool for gait analysis in unilateral TT amputees, and the results are comparable with those obtained using a Vicon gait measurement system.