Decoding a Neurofeedback-Modulated Cognitive Arousal State to Investigate Performance Regulation by the Yerkes-Dodson Law

Enhancing the productivity of humans by regulating arousal during cognitive tasks is a challenging topic in psychology that has a great potential to transform workplaces for increased productivity and educational systems for enhanced performance. In this study, we assess the feasibility of using the Yerkes-Dodson law from psychology to improve performance during a working memory experiment. We employ a Bayesian filtering approach to track cognitive arousal and performance. In particular, by utilizing skin conductance signal recorded during a working memory experiment in the presence of music, we decode a cognitive arousal state. This is done by considering the rate of neural impulse occurrences and their amplitudes as observations for the arousal model. Similarly, we decode a performance state using the number of correct and incorrect responses, and the reaction time as binary and continuous behavioral observations, respectively. We estimate the arousal and performance states within an expectation-maximization framework. Thereafter, we design an arousal-performance model on the basis of the Yerkes-Dodson law and estimate the model parameters via regression analysis. In this experiment musical neurofeedback was used to modulate cognitive arousal. Our investigations indicate that music can be used as a mode of actuation to influence arousal and enhance the cognitive performance during working memory tasks. Our findings can have a significant impact on designing future smart workplaces and online educational systems.