Threshold-Based Machine Learning Model for Hand State Classification Using EMG Sensors and Support Vector Machines

Electromyography sensors (EMG) are techniques used to get the electrical signals generated by muscle activation, providing valuable insights into muscle activity and its relation to the nervous system. This chapter addresses the challenge of achieving accurate, real-time classification of hand states using a hybrid system combining threshold-based detection and machine learning algorithms. The proposed method focuses on the development of an accurate and efficient muscle activation detection system that utilizes the EMG sensors along with Support Vector Machines (SVM) for real-time classification of two hand states. The key contribution of this research is to incorporate a robust architecture that is adaptable to diverse muscle activity patterns and has the potential for deployment in real-time applications such as rehabilitation centers and prosthetics. In this chapter, the model is trained on two users, with varying muscle strength, and evaluations are done using accuracy, precision, and confusion matrix. Testing showcased high accuracy and precision.