Virtual reality rehabilitation in spinal cord injury patients

Villiger, Michael (2012) Virtual reality rehabilitation in spinal cord injury patients. PhD thesis, ETH Zurich.

Full text not available from this repository.


Patients with an incomplete spinal cord injury (iSCI) are limited in their lower limb motor function and are also associated with a wide range of physical and psychosocial problems, including in particular neuropathic pain. The consequences of spinal cord injury are challenging to treat and the potential for recovery is still very limited. Although recent trainingbased neurorehabilitation approaches using virtual reality (VR) show some promise, the methods employed to date address only patient motivation issues and thus do not tap into the potential of VR. In particular, they do not directly promote activation of the underlying sensorimotor network to (re)activate cortical structures involved in control of the lower limbs. This thesis makes a contribution to the understanding of the brain processing involved in lower limb movements mediated by VR, and its application to the treatment of iSCI patients. It is known that observing, imagining, and even understanding motor actions activate the neural networks involved in motor execution. Using VR training to activate this supraspinal network projecting to surviving parts of the spinal tract may thus promote rearrangement of the network and bring substantial locomotion benefits and pain reduction to iSCI patients. To test this principle, a new VR neurorehabilitation system was developed to provide taskspecific VR training of isolated movements. The system displayed life-sized virtual lower limbs on a large screen, controlled by the user via size-adjustable shoes with integrated orientation sensors. This thesis contains four studies, all based on the combination of observation and motor imagery, i.e. online motor imagination, a cognitive process in which a subject imagines himself/herself in the displayed movement situation. The first study using fMRI shows that it is possible to extensively activate the neural correlates of the lower limb motor execution network using online imagination, even in the absence of overt movement. The first ever VR training system for lower limbs combining action observation, imagination and execution was developed, with four training applications for different lower limb muscles and functions. In the second study, a first investigation of the system in a pilot single-case study on iSCI patients showed that the training was well accepted by the patients and seemed to be enjoyable even if they were not familiar with VR. The motivation was high, which is promising for simple training of repetitive movements. The VR rehabilitation system required active patient effort at all times and the results demonstrated improvement of lower limb motor functions and reduction of pain intensity. The third study, a training study on iSCI patients with motor dysfunction and neuropathic pain, revealed positive effects on a short-term and longitudinal development. In the fourth study, the findings were extended for a VR lower limb training scenario called “footbag”, showing that motor online imagination and imitation for the VR “footbag” game and those of playing the game without or with reward revealed similarly activated brain regions. Furthermore, the results suggested that playing a game and being in control of the VR lower limbs is a rewarding activity per se. The findings of this thesis extend previous results promoting the use of pure motor imagery, or of action imitation, as therapeutic approaches in lower limb neurorehabilitation. The use of VR techniques in neurorehabilitation may have profound implications on the facilitation of post-injury retraining of function and potentially promote reparative plasticity and functional recovery. It has been shown that the VR system improves motor function and reduces neuropathic pain in iSCI patients. A further advantage of the system is its low cost compared to other training technologies, e.g. robotics. Essentially, the patients were able to generalize the training of individual movements to achieve functional improvements on their own with no task-specific training. An open question for future investigation is the extent to which these results can be generalized to other neurological patient groups, e.g. stroke.

Actions (login required)

View Item View Item