Improving Wheelchair Driving Skills through a Virtual Reality Simulator in Cerebral Palsy: an Evaluation of Performance Metrics

Powered wheelchairs (PWs) significantly enhance the quality of life for individuals with neuromotor disabilities. Nevertheless, for most people who use a powered wheelchair, the training is exhausting. Moreover, to date, there are no measures for assessing driving skills other than the WST (Wheelchair skill test) and PMP (Powered Mobility Program). Virtual reality (VR) emerges as a promising alternative for PW training in response to the extensive training demands of therapists. This Thesis aims to develop a VR simulator to enhance PW driving skills in children with cerebral palsy (CP). The VR application mirrors tasks from the powered mobility program (PMP), assessing PW driving ability, with fully immersive and semi-immersive modalities for those experiencing cybersickness. To ensure inclusivity for CP children unable to control VR joysticks, the system integrates various access technologies (ATs). An existing limitation of VR systems is the lack of online performance quantification compared to standardized clinical tests such as the PMP. Hence, one objective of this thesis is to provide an objective quantification of performance. The VR simulator was evaluated with 19 able-bodied adults and 4 children with CP. Among the adult group, 3 cases of cybersickness were reported, while none were reported among children with CP. After conducting literature research to identify widely used metrics, two methodologies for assessing driving skills were developed and tested. Comparisons were made between the calculated scores in VR and the evaluations from the PMP. Through data analysis, a methodology for evaluating VR performance demonstrated a statistical correlation with PMP evaluations conducted by clinicians during VR testing. This thesis lays the foundation for objectively evaluating driving skills by comparing them with subjective measurements provided by clinicians.