Using ANN on EEG signals to predict working memory task response

Many authors have shown that performing working-memory tasks causes an elevated neuronal activity in several areas of the human brain, which suggests information exchange between them. Since the information exchanged, encoded in brain waves is measurable by electroencephalography (EEG) it is reasonable to assume that it can be extracted with an appropriate method. In this paper we present a method for extracting the information using an artificial neural network (ANN), which we consider as a stimulusresponse model. The EEG was recorded from three subjects while they performed a modified Sternberg task that required them to respond to each trial with the answer "true" or "false". The study revealed that a stimulus-response model can successfully be identified by observing phase-demodulated theta-band EEG signals 1 second prior to a subject's answer. The results showed that the model was able to predict the answers from the EEG signals with an average reliability of 75% for all three subjects. From this we concluded that stimulus-response model successfully observes the system states and consequently predicts the correct answer using the EEG signals as inputs.