Development of Assistive Platforms for the Disabled using Brain Signatures

Abstract

Many people suffer from the severe neuromuscular disorders. This leads them inefficient to interact with the outside world or even to do their daily routine work. These disabilities can be overcome with the help of assistive devices. Assistive devices can be controlled using brain signals corresponding to particular imagination. The neural signals are processed to generate the control commands for activating the assistive devices for particular mental activity. Nowadays number of assistive devices is available; however most of them are inefficient in accuracy or speed. This paper presents a simple and effective framework which interfaces assistive devices to the OpenViBE environment to improve the lives of physically disabled person. The motor restoration and speech impairment issues are addressed in this paper. Development of assistive platform is carried out in two steps; in first step the brain signals are decoded with high accuracy and in the second step these signals are converted into the control command to operate the assistive device according to the user's desire. Combination of AR coefficients, Hilbert transform and PSD based features is proposed. These features are classified by the SVM classifier to provide the accuracy of 88.5%. This in turn can control the assistive devices effectively.