After stroke, a patient will typically need to do many repetitions of specific exercises over a short period of time in order to regain muscle definition and strength in their affected upper limb. In this initial project a prototype system using a Wiimote camera and a bespoke ‘IR glove’ was developed to track a user’s hand and finger movements. These movements were then used as inputs to a number of computer games designed to encourage the arm and hand movements required to recover from stroke. (more) Further work has been undertaken by Ph.D. project students utilising Microsoft Kinect and then Kinect II on the Kinect II developer’s program, to move this project forward. It is hoped that the technologies will be able to detect ‘bad’ behaviours in patient’s rehabilitation exercises, and highlight them to the patient. By these means the system will hopefully be able to prevent patients from doing their exercises in an incorrect way, which could result in injury.