Player uncertainty in navigation games and its applications

Treatment of progressive dementia such as Alzheimer's disease is challenging partly due to delay in diagnosis. Here, I propose developing a model of the players' behaviour in navigation-based games to detect early stages of cognitive impairment. I will do so by adapting a Bayesian ideal-observer model based on visual simultaneous localisation and mapping. The model will fit and predict the player's moment-by-moment movement decisions given the first-person view and the map of the game environment. I will fit the model to the trajectory of normal players and Alzheimer's patients playing the game, which will yield parameters that indicate the levels of visual, motor, and memory noise of each individual player. I will then examine what combination of parameters can best differentiate patients from normal players. Early diagnosis would lead to early treatment, thereby improving the quality of life for patients. Moreover, the research has the potential to enhance gameplay experience by keeping the players engaged.