Functional Electrical Stimulation with Machine Learning for Stroke Prediction and Rehabilitation

This paper presents a comprehensive approach to brain stroke management, focusing on prediction and rehabilitation aspects. It develops a machine learning algorithm that predicts future strokes based on patient features like age, gender, BMI, sugar levels, pulse rate, and overall health condition. The second phase introduces a novel rehabilitation method using Functional Electrical Stimulation (FES), which mimics natural brain-to-muscle signals. The simulated FES therapy improves muscle coordination, especially for patients with paralysis or compromised brain-to muscle connections. The machine learning stroke prediction model shows reliable predictions and provides valuable insights into key features influencing stroke likelihood. The FES system simulation shows promising outcomes in rehabilitating the prosthetic arm, enhancing patient coordination and adjustable degrees of muscle movement. This interdisciplinary project bridges the gap between predictive analytics and rehabilitation technology, offering a holistic solution for stroke management. The integration of machine learning and FES presents a promising avenue for personalised stroke care, with potential applications in real world healthcare settings.