Recognition of Healthy Individuals’ Perceptual Decisions Using Electroencephalogram Signals and Artificial Neural Networks

Abstract

Purpose: The process of making a decision based on available sensory information is called “Perceptual Decision Making”. The manner in which this decision is made has a direct impact on a person's social and personal relationships. Despite numerous studies in the field of perceptual decision making, there is still no robust system that can recognize people's perceptual decisions objectively. To this aim, this study aims to examine the relationship between EEG signals and perceptual decision making in healthy individuals.

Materials and Methods: The research employs an online EEG dataset based on visual stimuli, including faces and cars, obtained from 16 participants. Since there is no binary decision-making mode in the brain and there is an uncertainty in which each option has a special weight in decision-making and finally the option that passes a threshold is selected, this research has tried to incorporate this uncertainty into the final model to improve perceptual decision recognition system performance. For this purpose, a fuzzy radial basis function (FRBF) network was utilized.

Results: After extracting 26 features from the preprocessed EEG signals, Friedman’s non-parametric statistical analysis was performed, revealing that differences in the coherence of stimulus representations have a greater impact on an individual's decision-making process than spatial prioritization. Then, FRBF network classifier, with the extracted features from TP9 and TP10 channels as input, achieved an accuracy of 90.3% in classifying the test data as either a "face" or "car".

Conclusion: The classification accuracy results showed that the proposed method is an effective procedure for recognition of human decisions.