Evaluation on game concentration with multi-scale fuzzy entropy based on EEG signals

Abstract

Concentration is the index to measure a person concentrating on certain thing or activity. Finding a valid physiological characteristic index to represent concentration level is important for attention training and experience improvement. In this study, the multi-scale fuzzy entropy (MSFEn) was proposed as an EEG feature to evaluate concentration level. Firstly, a game platform was designed independently to arouse the players’ concentration during the game, and the physical characteristics index Rect_y was defined to directly represent the players’ emotional concentration. Secondly, a portable single-channel EEG device was adopted to collect EEG signals of players during the experiment. Hilbert–Huang transform, the MSFEn and Grey correlation analysis were used for signal processing. The behavioral performances of participants were investigated with a questionnaire after each participant playing game. The statistical results showed that the MSFEn maintained a high correlation with the Rect_y at different game speeds and different scale factors for each player, with the median of about 70%. Compared with conditional entropies, the MSFEn can better reflect relatively short-term concentration and long-term overall game concentration of participants during the game.