An automatic detection and localization of P300 sub-components using ICA

Independent component analysis (ICA) has been applied for separation (BSS), detection and localization of P300 sub-components in this project. Event-related potential (ERP) is obtained from EEG data of the schizophrenia patients. The frequency analysis method is applied to extract P300 sub-components from the background EEG signal that contains noise and artefacts. By using JADE algorithm and correlation analysis, P300 sub-components (P3a/P3b) have been separated effectively. The components are then localized based on the correlations between the electrodes and the estimated source signals.