Elements: clinical design and evaluation of a virtual reality augmented workspace for upper-limb rehabilitation of traumatic brain injury

摘要

Traumatic Brain Injury (TBI) refers to cerebral damage caused by external physical force, and results in a range of cognitive, and physical impairments. Accordingly, developing new technologies to further TBI rehabilitation is a central focus of research. One technology that offers significant advantages is Virtual Reality (VR). This thesis describes the design and initial testing of an upper-limb VR-rehabilitation program for TBI (called Elements). The aims with the Elements system were to create virtual workspaces that were theoretically sound, challenging, and engaging, yet could be tailored to participants’ individual needs. The system has both rehabilitation and assessment functions. The rehabilitation package consists of two sets of virtual environments (VEs) viewed on a 1020 mm (44 in.) horizontal LCD monitor. The four goal-based VEs utilise a stimulus-response format, where participants move real objects to cued locations on the screen. The difficulty of these VEs is scaled to place greater requirements on motor planning by varying task constraints (e.g. randomising the presentation order of movement cues). Additional visual and auditory movement feedback is provided in these VEs to facilitate functional movement. In contrast, the three exploratory VEs have no clear ‘goal’. Here participants freely (even creatively) interact with these VEs, which encourages them to devise and execute their own motor plans. For assessment the system automatically tracks movement accuracy, speed, and efficiency during the goal-based VEs. Feedback plots from this data are used to provide participants with knowledge of their results. Participants underwent 12 one-hour sessions of VR-rehabilitation over 4 weeks. Two empirical trials were conducted to assess the system. Study 1 was a multiple case-study (with 3 participants), and applied an ABA design. Participants were assessed on the system-measured variables and tests of unimanual, and bimanual function over baseline and treatment phases. Participants improved on movement accuracy, efficiency and unimanual function over the treatment period. These improvements were largely maintained in the second baseline phase. Mixed improvement was seen on speed and bimanual coordination. Accordingly, based on these generally positive results, a further larger sample trial was conducted. Study 2 was a within-groups investigation. Our 9 participants’ upper-limb and neurobehavioural function were measured before and after one month of normal therapies alone, and following one month of normal and VR-rehabilitation. Participants demonstrated no significant improvements over the normal rehabilitation period (except for the speed variable). Statistically significant improvements in movement accuracy, speed, efficiency, general upper-limb, and neurobehavioural function (especially memory/attention) followed VR-training. Finally, a discussion on these findings and their implication for the VR-rehabilitation field, and the application of the ITE model in system design is presented. Possible areas for future research (e.g. inclusion of other patient groups, use of brain imaging technology) are also outlined. It was concluded that the results of these trials provide good initial support for the Elements system, and justify further larger sample studies. These are the necessary next steps in trialling the system, and supporting VR’s application in TBI rehabilitation.