Development of EMG Controlled Electric Wheelchair Using SVM and kNN Classifier for SCI Patients

Spinal cord injury (SCI) is a devastating condition which can lead to quadriplegia or paraplegia depending upon the level at which injury has occurred resulting in restricted mobility and reduced quality of life. Electromyogram (EMG) signal based human machine interface (HMI) is a system that can be used as an assistive technology to enhance the life of SCI patients. The present paper aims to provide an assistive device (electric wheelchair) for patients suffering from SCI at lower cervical level. EMG signal is used to control the movement of the electric wheelchair, for which time domain feature are used to train support vector machine (SVM) and k-nearest neighbour (kNN) classifier in python 3.5 software to categorize the 5 different movement controls. Raspberry pi 3 is used to acquire signals from MyoWare sensor and processes these signals to provide control pulses to DC motor drive, which finally executes by the electric wheelchair.