Machine Learning-Based Analysis of Lower Limb Activity For Human Movement Understanding

Electromyography (EMG) signals, when connected to the lower limb, offer valuable insight into muscle activity, movement patterns, and biomechanical performance. The application of machine learning (ML) techniques to lower limb EMG signals is summarized in this article, with a focus on the importance of these methods in a variety of domains, such as biomechanics, sports science, and rehabilitation.ML algorithms applied to lower limb EMG data enable the estimation of muscle activation patterns, recruitment dynamics, fatigue levels, and synergistic muscle coordination during movement tasks. Furthermore, ML techniques facilitate the estimation of biomechanical parameters such as joint angles, joint moments, and muscle forces, providing a comprehensive understanding of lower limb function and performance. In this study, performance metrics for five algorithms—SVM, KNN, a hybrid of SVM and KNN, RF, and MLP—were evaluated, demonstrating the effectiveness of the hybrid model in achieving the highest accuracy. As ML continues to evolve, its integration with lower limb EMG promises to further enhance our understanding of human movement and contribute to the development of innovative interventions aimed at improving mobility, function, and quality of life.