A Bayesian Graph Neural Network for EEG Classification — A Win-Win on Performance and Interpretability

With the deepening of neuroscience research, data mining of brain signals is becoming an emerging topic. Among various brain signals, electroencephalography (EEG) has attracted more and more attention due to its advantages of non-invasiveness, portability, and low cost. EEG modeling and analysis play a vital role in human healthcare. Although many machine learning algorithms have been successfully applied to data mining of EEG signals, few of them achieve a win-win in classification performance and interpretability. In this paper, we propose a Bayesian graph neural network named BayesEEGNet. Considering an electrical impulse between two nodes in the brain as a Poisson process, the countless electrical impulses generated by the brain in a period are represented as an infinite number of connection probability graphs. After coupling and transforming these probability graphs, we interpret the brain’s electrical activity state as the brain’s perceptual state. Benefiting from the joint optimization of Bayesian modules and deep neural networks, our model shows superior classification performance in sleep stage classification and emotion recognition tasks. Meanwhile, our model is able to learn interpretable functional connectivity relationships between EEG channels without any prior knowledge.

I. Introduction

With the advancement of neuroscience, brain signal data can be collected at a lower cost and with a higher signal-to-noise ratio. The data mining of brain signals has also received more and more attention from researchers. First, data mining of brain signals plays an important reference role for disease diagnosis (sleep disorders, mental illness, etc.) [1], [2]. Second, mining the potential connections between multi-channel brain signals has momentous implications for our understanding of brain function. Last but not least, mining brain signals is also a vital part of realizing brain-computer interface.