Pareto-based Dynamic Difficulty Adjustment of a competitive exergame for arm rehabilitation

Although Dynamic Difficulty Adjustment (DDA) in rehabilitation-based exergames features the conflicting objectives of intensity and performance, traditional DDA methods have only addressed this problem in terms of single or weighted objectives. Deviating from previous works, this work presents a Pareto-based DDA (PDDA) for a Competitive rehabilitation-based exergame which is capable of generating a set of trade-off solutions between the two conflicting objectives. Consequently, based on the rehabilitative needs, a unique solution is selected via the trade-off worth information of the generated solutions to modify the game parameters or difficulty. In order to evaluate the performance of the proposed PDDA, two experimental studies were conducted based on 12 and 10 unimpaired participants respectively. The first study which was aimed at evaluating the performance of the proposed PDDA in terms of the in-game parameters shows that the proposed PDDA is capable of generating multiple trade-off solutions as well as effectively adjust the game parameters or difficulty. The second study which was aimed at evaluating the proposed PDDA in term of user experience shows that PDDA demonstrate statistical difference with a random DDA method in terms of the Intrinsic Motivation Inventory (IMI) and the Flow Experience Measure (FEM) metrics. Since rehabilitation outcomes generally rely on both intensity and performance, the ability of the proposed PDDA to increase effort, enjoyment, competence, and flow will help to greatly facilitate frequent and intense exercise as well as performance in rehabilitation-based exergame.