Identify Non-Fatigue State To Fatigue State Using Causality Measure During Game Play

In this paper, Granger causality (GC) and New causality (NC) analysis methods are applied in frequency domain to reveal causality changes from non-fatigue state to fatigue state with EEG signals during video game-playing. EEG signals were recorded while a subject was playing video-games. Results show that fatiguing phenomenon was observed in 15 subjects using NC in [20, 30] Hz while only 13 subjects were identified with GC for comparison. The NC further showed the bi-directional causality changes between the two hemispheres during unilateral forearm movements. We noticed that half of the subjects had predominant active hemisphere while the other half showed the opposite, especially to the ones with higher fatigue level. The findings demonstrate that the NC method is better than the GC to reveal causal influence between homologous motor areas of active and inactive hemispheres in this study.