Validation of Response by Classifying EEG Gathered from Different Areas of the Brain

Abstract

This paper uses EEG data collected while conducting a visual search experiment where 31 participants were asked to look for a target in a pool of 63 distractors. Each participant was provided with 300 different trials and in each trial the target was hidden in a randomly chosen position. In this paper artificial neural network (ANN) and long short-term memory (LSTM) models are applied for recognition and classification of electroencephalographic (EEG) patterns associated with correct or incorrect detection of targets. This paper provides a comparison between EEG signals captured from different lobes, or areas of the brain [pre-frontal (Fp), frontal (F), temporal (T), parietal (P), occipital (O), and central (C)] to see which can provide more accurate prediction of a participant's response for target in a visual search experiment. It has been showed that the data taken from a certain portion of the brain improved the LSTM model and ANN Model by more accurately predicting if a participant has provided correct or incorrect response. In future this can help us create BCI (Brain Computer interface) experiments where data taken from specific areas of the brain can help track a participant's performance in a task. In BCI experiments the speed of processing data is of utmost importance and if our models can predict a participant's performance in a shorter time, they can help us improve BCI undertakings.