Kinesthetic Motor Imagery Based Brain-Computer Interface for Power Wheelchair Manoeuvring

Abstract

Patients who are suffering from diseases like motor neurone diseases (MND), or trauma such as spinal cord injury (SCI), and amputation is not able to move. This paper presents work on combining the power wheelchair designed to aid the movement of the disabled patient, and a Brain-Computer Interface can be used to replace conventional joystick so that it can be controlled without using hands. The brain signal emanated during Motor Imagery tasks can be converted into control signal for power wheelchair manoeuvring. In this research, five subjects are requested to perform six Kinesthetic Motor Imagery tasks, and Electroencephalography (EEG) signals are recorded. The elliptic filter was used to remove power line noise. Three features, namely Fractal dimension (FD), Mel-frequency Cepstral Coefficients (MFCCs) and a combined feature of FD with MFCCs were extracted and evaluated by using Multilayer Perceptron Neural Network (MLPNN). The Levenberg-Marquardt training algorithm is used to train the networks, and the classification result of the MLPNN using a combined feature of FD with MFCCs achieved an average accuracy of 91.7%. The developed model is tested and evaluated with the simulated virtual environment created by MATLAB graphical user interface (GUI). The result suggests that the combined feature of FD with MFCCs and MLPNN can be used to classify Motor Imagery signal for directional control of powered wheelchair.