Characterizing Neural Activity During Video Game Engagement Using EEG Sensor-Based Topological Dynamics Analysis
Kumar Gaurav, Jaykumar Landge, Tharun Kumar Reddy
Abstract
Understanding the complex neural activity patterns in the developing brain is a major neuroscience challenge. Characterizing neural activity during video game interaction allows for naturalistic and engaging brain dynamics research. Electroencephalography (EEG) sensors capture neural signals with high temporal resolution noninvasively, making them ideal for studying real-time brain activity during dynamic tasks like gaming. Understanding the complex dynamical phenomena of EEG patterns corresponding to different neural activities requires nonlinear analysis and feature extraction. Phase space reconstruction (PSR) is a typical nonlinear technique used to reveal the dynamics of the brain's neural system. Recently, topological signal processing (TSP) has been utilized to explore the properties of phase space, providing a powerful tool for analyzing the phase space. In this letter, a novel topological EEG nonlinear dynamics analysis approach is proposed using PSR techniques to convert EEG sensor signals into phase space, with the persistent homology tool employed to explore the topological properties of phase space. The proposed approach is implemented and experimentally validated on a publicly available EEG-based mobile brain–body imaging dataset. The framework demonstrates its efficacy in characterizing neural activity during game interaction with average accuracy of 90.31% for model 1 (TP9 and TP10) and 93.49% for model 2 (all four channels), outperforming existing deep learning approaches. This letter represents the first investigation into neural activity characterization-oriented EEG topological feature analysis, providing novel insights into the nonlinear dynamics analysis and feature extraction of the brain's neural system.