BRAIN ACTIVITIES FOR MOTOR MOVEMENT

Brain Computer Interface (BCI) is hardware and software system which allows interaction between the human’s brain and some surrounding activities without depending on their muscles or peripheral nerves. The main objectives of this project are to design a brain computer interface algorithm that takes Electroencephalography(EEG) signals as its input, translates them into commands for movement control and to test the performance of the designed algorithm on human subjects.
The research covers the procedure of designing the BCI algorithm and this consists of three stages firstly recording EEG brain signals, secondly EEG signals pre-processing, Last stage is EEG signals classification.
The EEG signals classification is divided into 2 parts which includes feature extraction and feature classification. Multivariate adaptive auto regressive (MVAAR) method is used in the feature extraction part because it is suitable for motor imaginary. Feature vectors are used to differentiate the different brain activity signals associated with the user’s attention, Linear Discriminate Linear (LDA) method is used in feature classification step to achieve these goals.
The Feature extraction method MVAAR couldn’t extract the actual feature for the four movements so we couldn’t classify between them.