Auditory brain response detection using a portable EEG headset

In general terms, evoked potentials are electrical signals generated by the nervous system in response to a stimulus. Auditory evoked potentials (AEPs) are generated in response to an acoustic stimulus. Measuring the electrical response of auditory system gives many information about the status of individuals hearing. Auditory brain response (ABR) is an AEP and can be detected using EEG technology and signal processing techniques. In this thesis, an ABR detection system has been implemented. Experiment procedure was designed using auditory oddball paradigm. An acoustic stimulus has been sent to the subject and a marker about the stimulus has been sent to the recording software simultaneously while recording EEG. A low cost, wireless EEG headset was used to record EEG data under auditory stimulus from 13 subjects. Raw EEG data has been processed by using epoch extraction, event related potential (ERP) averaging, and independent component analysis (ICA) methods. Some features were extracted about the auditory stimulus. Then the extracted features were used to classify the data to understand if hearing has occurred or not under given stimulation. Results of the experiments showed that the implemented ABR detection system detected the sound and silence stimulation with 85% accuracy.