Multimodal Fusion of EEG and Eye Data for Attention Classification using Machine Learning

The cognitive process of attention can be classified into two broad categories: Internally Directed Cognition (IDC) and Externally Directed Cognition (EDC). Classification of attention state based on goal-oriented tasks can assist in the diagnosis of mental disorders, such as attention deficit hyperactivity disorder (ADHD), schizophrenia or obsessive-compulsive disorder (OCD). Certain time-sensitive neurophysiological characteristics such as electroencephalogram (EEG) and eye movement can serve as the basis for differentiating between the two attention states. This paper introduces an efficient binary classification model for distinguishing between IDC and EDC. After preprocessing and analysis of captured data, a multimodal machine learning model is employed for classification of attention. The model achieved an accuracy of 74.64% outperforming the state-of-the-art model.