VR-DLR: A Serious Game of Somatosensory Driving Applied to Limb Rehabilitation Training

Abstract

Hemiplegia is one of the common symptoms of stroke, especially the motor dysfunction of upper limb has a great impact on the patients' daily life, and it is also one of the difficult problems in rehabilitation treatment. However, traditional rehabilitation training lacks pleasure and easy to make patients shrink. We use virtual reality technology to design VR-DLR (VR driving for limb rehabilitation), a serious game of somatosensory driving that combines hardware, software and integrates sight, hearing and tactile sensations. We design parameterized steering wheel training actions according to the level of the patients' limb disorders and the rehabilitation physician's treatment plan. The patients' training task is to control the virtual car to collide more coins and to avoid fewer obstacles. We collect and record the patients' interactive operation data, and design evaluated model to evaluate the patients' score and rehabilitation effect. VR-DLR supports two modes of fixed speed and free driving and has adapting difficulty in each mode according the limb disorders of patients. Moreover, in order to meet the needs of different patients, VR-DLR has two display modes of VR headset and 3D annular projection screen. A user study shows that VR-DLR can significantly increase the interest, initiative, confidence of patients in rehabilitation training compared with traditional training methods.