Thought controlled IRCC using cross-correlation of different frequency band sequence

In recent years, researchers have developed a wide range of Brain Machine Interface applications to aid differentially enabled communities, which includes speech communication and wheelchair navigation. Intelligent Robot Chair with communication aid (IRCC) involves classification of navigational tasks (FORWARD, LEFT, RIGHT & RELAX) along with tasks related to speech (YES, NO and Help), to provide an alternative navigation system with a communication aid for the DE patients. The navigational tasks and the binary YES/NO tasks can be used for navigating a wheelchair and also for speech communications using any oddball paradigm. The additional tasks ‘HELP’ can be used to help the patient during an emergency. The proposed system records the 8-channel wireless electroencephalography signal from twenty subjects while the subject performs the seven different tasks. The recorded brain wave signals are preprocessed to remove the interference waveforms and segmented into six frequency band signals, i.e. Delta, Theta, Alpha, Beta, Gamma1-1 and Gamma-2. The frequency band signals are segmented into frame samples of equal length and are used to extract the features using cross-correlation (r) function. This study presents statistical features of cross-correlation based distinctive feature extraction technique. The cross-correlation features are extracted from combinations of different frequency band signals. The extracted feature sets are tenfold cross validated using multi-class support vector machine using RBF kernel algorithm. From the results, it is observed that the Mean feature set has the maximum classification accuracy of 88.65 % and standard deviation feature has the minimum classification accuracy of 78.51 %.