Maintaining patient engagement is a key challenge in any rehabilitation paradigm. In conventional hospital-centric therapy, the therapist ensures patient engagement and motivation towards therapy. In robotic rehabilitation, this human chaperone is absent. We are trying quantify the patient mental engagement levels as they perform therapy and adapt the system accordingly. We are using EEG, EMG and the patient's interaction with the robot itself as parameters for this quantification. In order to make therapy engaging, we have developed a set of games to simulate therapy exercises. The system is targeted towards stroke patients. The current implementation involves a straight line following game, a handwriting simulator and an eating task (see figures below).
The system is also being utilized for rehabilitation of children suffering from traumatic brain injury.