Development and clinical trial of virtual reality-based cognitive assessment in people with stroke: preliminary study.

Stroke and traumatic brain injury affect an increasing number of people, many of whom retain permanent damage in cognitive functions. Conventionally, cognitive function has been assessed by a paper-based neuropsychological evaluation. However these test environments differ substantially from everyday life. This problem can be overcome by using virtual reality (VR) to objectively evaluate behaviors and cognitive function in simulated daily activities. With our virtual shopping simulation, we compared people who had undergone a stroke with control participants in an immersive VR program that used a head-mounted display (HMD). We evaluated user satisfaction with the tests, complications, and the user interface. Significant differences were consistently found between the stroke group and the control group for the following tasks: stage 1 performance index, interaction error; stage 2 delayed recognition memory score, attention index; and stage 3 executive index (p < 0.001). Perceptive dysfunction, visuospatial dysfunction, level of computer experience, and level of education affected the performance of the stroke group. The frequency of complications in the stroke group, calculated using the cut-off score for the Simulator Sickness Questionnaire, was 9.6% for nausea, 41.9% for oculomotor complications, and 25.8% for disorientation. The frequency of complications between the stroke and control groups was not significantly different. Thirty-five percent of participants in the stroke group and 13% in the control group reported difficulties with using the joystick. This computer-generated VR-based cognitive test shows promise in assessing cognitive function in patients with stroke. More refinements are needed in the user interface and the projection methods.