Analysis of Electroencephalography (EEG) Signals and Its Experimental Design

Brain-Computer Interface (BCI) that interprets human brain signals to control computers and different communication devices comprises a complete system including the software and hardware. This research work focuses on the experimental design framework of the electroencephalograph (EEG) signals. EEG brain signals of the subject are acquired whilst performing five different tasks by using Emotiv Epoc+ device. For preprocessing, a completely automated method for detection of artefactual Independent Component ICs from EEG data using automatic EEG artifact detector based on the joint use of spatial and temporal features (ADJUST) have been used. A Matlab tool is developed to load the raw data of Emotiv into the EEGLAB for filtering and basic feature extraction, such as Power Spectral Density (PSD) and Hjorth parameter. The goal is to develop a pattern recognition algorithm performing hinge joint movement that provide full range of motion (ROM) from extension to flexion that will be used for post stroke rehabilitation. The results show that beta frequency band differs for every task performed during experiment.