Exploration of an EEG-Based Cognitively Adaptive Training System in Virtual Reality

Virtual Reality (VR) is effective in various training scenarios across multiple domains, such as education, health and defense. However, most of those applications are not adaptive to the real-time cognitive or subjectively experienced load placed on the trainee. In this paper, we explore a cognitively adaptive training system based on real-time measurement of task related alpha activity in the brain. This measurement was made by a 32-channel mobile Electroencephalography (EEG) system, and was used to adapt the task difficulty to an ideal level which challenged our participants, and thus theoretically induces the best level of performance gains as a result of training. Our system required participants to select target objects in VR and the complexity of the task adapted to the alpha activity in the brain. A total of 14 participants undertook our training and completed 20 levels of increasing complexity. Our study identified significant differences in brain activity in response to increasing levels of task complexity, but response time did not alter as a function of task difficulty. Collectively, we interpret this to indicate the brain's ability to compensate for higher task load without affecting behaviourally measured visuomotor performance.