Development In Limb Prosthetics Using Machine Learning Algorithms

Abstract

Machine learning is revolutionizing the healthcare industry by enabling real-time prediction of ortosis needs and enhancing prosthetic selection, training, fall detection, and socket temperature control. Artificial limbs can help people with disabilities to regain their ability to move. An important method in artificial limb technology is pattern identification of limb movement intention.

This paper highlights how machine learning algorithms play an important role in upper and lower limb amputation.

For upper limb EEG and EMG processing enhances classification accuracy in prosthetics, enhancing upper limb movement control performance and control performance in paraplegics with elbow replacements, demonstrating an 8.9% improvement ratio in real-time applications.

For lower limb prosthetics we implemented various algorithms namely KNN, SVM, and QDA, Random Forest, Random Forest with estimator 100, and Random Forest with activator 60 to test the best accuracy.