Abstract
Virtual reality (VR) has been used as a technological medium to deliver a broad range of therapeutic rehabilitation interventions over the last 30 years. These can range from the treatment of specific exposure phobias to neurological rehabilitation such as phantom limb pain and stroke rehabilitation. For the latter, mirror therapy has been shown to be an effective treatment for one sided, upper limb rehabilitation following stroke. Mirror therapy works by fooling what the brain is seeing, in the case of stroke, that the affected limb is carrying out movements, to promote positive neurological re-wiring of the brain which can lead to relearning of lost motor control skills. However, mirror therapy does not work for everyone (deemed ``non-responders") and there exists room for immersive VR to further assist the ``fooling'' of the brain and the possible effect this could have on rehabilitation.
Immersive VR utilises head-mounted displays to surround the user in a virtual environment which can be used to carry out virtual rehabilitation. In the case of mirror therapy, we can present people with stroke with the ability to carry out virtual mirror therapy. By utilising immersive VR hardware, we can address some of the limitations of conventional mirror therapy as it is carried out in clinical environments such as a observing the mirrored limb in a spatially congruent position, decoupling the user from their real environment and their actual hands, thus, perhaps even providing a more convincing ``fooling’’ of the brain that their affected limb is carrying out movements.
In this thesis, we report on the design, implementation and evaluation of an immersive VR mirror therapy stroke rehabilitation system (BeST-ART VR). We approach this from an ecological perspective: we aim to deliver not just any implementation of mirror therapy, but mirror therapy as carried out in real clinics. By doing so, our developed system allows people with stroke the ability to carry out the same rehabilitation exercises (protocol) that they would experience in a clinical rehabilitation in our developed immersive virtual environment. We identified and designed our system based around one such protocol, BeST, that is used in daily practice at rehabilitation clinics in Germany. The BeST protocol provides a validated, structured and standardised guide for clinicians on how to carry out conventional mirror therapy with people with stroke. We adapt this protocol for use in immersive environments and developed our immersive VR system to carry out mirror therapy as stipulated by the adapted BeST protocol, thus, providing our system with the ecological validity that other developed systems do not possess.
In mirror therapy, the person with stroke must suspend some disbelief and believe the mirrored limb they observe is their affected limb carrying out movements. In other words, they need to take ownership of that mirrored limb and the same is required for an immersive VR implementation of mirror therapy. The virtual hands need to be believable, that the user is in control of those hands and that they are in the correct spatial positions in the virtual environment. These aspects are forming the feeling of embodiment which we investigate in our first experiment. In our first experiment, we measure perceived embodiment towards a mirrored virtual hand model to show that high levels of embodiment can be achieved amongst healthy participants in an immersive neurorehabilitation (mirrored hand) scenario. By showing that high levels of embodiment can be achieved in a mirrored virtual hand, we can ethically use the system with people with stroke as it invokes the desired mirror therapy illusion effect.
In our second experiment, we adapt this system for clinical rehabilitation and run a clinical feasibility study at a rehabilitation clinic in Berlin, Germany. Our results show that it is feasible for clinicians to carry out their established, used in daily practice, mirror therapy protocol using our developed system. By following this protocol, we found that a majority of participants with stroke also experienced psychophysical effects during the intervention which could be the result of immersive VR providing a more convincing mirror therapy illusion. By showing that our system is clinically feasible, we can target home rehabilitation as the people with stroke will be carrying out the same hand exercises on their own and without a clinician present.
In our third experiment, we adapt the clinical system for home rehabilitation. We present the results of a home feasibility study in which people with stroke placed the system in their homes for one month and followed the same mirror therapy protocol without a clinician being present to operate the system and carried out our implementation of self-guided, autonomous rehabilitation. Our explorative feasibility study showed that people with stroke can operate our developed system to follow a mirror therapy protocol on their own. We gathered participant feedback from semi-structured interviews and carried out a thematic analysis based on their experiences using the system and having it in their homes. Four themes were derived based on the participant feedback: usability, efficacy, barriers to adherence and facilitators for adherence. Participants enjoyed many aspects of the immersive VR rehabilitation including gamification elements, ability to schedule their rehabilitation around their lives and perceived benefits to their affected limb.
Through the findings of these experiments, a consistent overall theme emerges showcasing the potential benefits of immersive VR for neurorehabilitation, in particular, mirror therapy. We argue that perhaps the time is now for researchers/clinicians to incorporate immersive VR hardware to target home rehabilitation for motivated and able people with stroke as an adjunct therapy after being discharged from the clinic. While we specifically target mirror therapy, our approach, methodology and findings can be applied to the rehabilitation of many other medical conditions. This work is of interest to the Human-Computer Interaction and Physiotherapy communities. Implications are particularly relevant for researchers and clinicians looking to incorporate immersive VR rehabilitation with their target cohort.