BMI using spectral energy entropy for colour visual tasks

Abstract

EEG signals are the electrophysiological measures of brain function and it is used to develop a Brain machine Interface. A Brain machine Interface (BMI) system is used to provide a communication and control technology for the people having severe neuromuscular disorders such as amyotrophic lateral sclerosis, brainstem stroke, quadriplegics and spinal cord injury. In this paper, a simple BMI system based on EEG signal emanated while visualizing of different colours has been proposed. The proposed BMI uses the color visual tasks and aims to provide a communication through brain activated control signal for a system from which the required task operation can be performed to accomplish the needs of the physically retarded community. The ability of an individual to control his EEG through the colour visualization enables him to control devices. Using spectral analysis, the alpha, beta and gamma band frequency spectrum features using energy entropy are obtained for each EEG signals. The extracted features are then associated to different control signals and a neural network model using probabilistic neural network (PNN) has been developed. The proposed method can be used to translate the colour visualization signals into control signals and used to control the movement of a mobile robot. The performance of the proposed algorithm has an average classification accuracy of 96.23%.