Difficulty adaptation in a competitive arm rehabilitation game using real-time control of arm electromyogram and respiration.

Abstract

Rehabilitation robots are often combined with serious games that motivate patients and keep them exercising at high intensities. A promising type of game are competitive rehabilitation games, but few difficulty adaptation algorithms have been presented for them. This paper thus presents the adaptation of difficulty in a competitive arm rehabilitation game based on two physiological signals: respiration and electromyography of the posterior deltoid. It consists of three smaller studies: an open-loop respiration study, a closed-loop respiration study (where a controller attempts to maintain respiration rate at preset levels), and a closed-loop electromyogram study (where a controller attempts to keep the electromyogram at preset levels). The studies control two difficulty parameters based on the physiological responses of one of the two exercising participants, though the ultimate goal is to control the physiological responses of both participants. Furthermore, all three studies are done with unimpaired participants. The closed-loop controllers achieved high correlation coefficients between desired and measured levels of respiration rate (r = 0.83) and electromyogram (r = 0.89), demonstrating that it is possible to control the physiological responses of unimpaired participants in a competitive arm rehabilitation game, thus controlling their level of workload and exercise intensity. In the future, the proposed method will be tested with patients undergoing rehabilitation.