Abstract
Mirror therapy is a non-invasive treatment procedure in which visual observation of a mirror-reflected moving limb ameliorates some sensorimotor symptoms associated with an impaired (or amputated) limb. Initially proposed as a possible treatment method for persistent phantom pain in limb amputees, over the years the therapy has also been applied in cases of limb paralysis or weakness due to stroke, complex regional pain syndrome, other neuropathic pain syndromes and sensorimotor malfunction. However, the precise mechanisms underlying its efficacy as a treatment are unknown, as are its neural underpinnings.
This thesis studied the neural correlates of mirror therapy. Specifically, it investigated whether the perception of hand movement through a mirror evokes activation in the primary motor cortex (M1) contralateral to the non-moving hand (the M1 ipsilateral to the moving hand). The primary hypothesis was the mirror visual feedback (MVF) of the hand movements in the mirror view conditions would increase ipsilateral M1 activation compared to the ipsilateral M1 activation evoked by the actual hand in the direct view conditions (where ipsilateral is defined with respect to the moving hand). Directly related to the main hypothesis, this study also investigated the influence of visual feedback (or observation of self-movement) on M1 activation.
A series of experiments were conducted using a unimanual hand movement and manipulations on visual observation. Electroencephalography (EEG) was used to record brain activity in healthy individuals and in people with upper limb amputation while they performed unimanual extension-flexion hand movement in different experimental conditions. The lateralized readiness potential (LRP) was measured in controls, and the steady-state movement related cortical potential (ssMRCP) was used as an alternative measure in amputees (given that LRP cannot be calculated); these were measured from continuous EEG as a marker of M1 activation associated with each experimental condition.
In Experiment 1, both the direct and mirror view of hand movements produced LRPs in healthy individuals, with the mirror view conditions revealing clear activation in M1 contralateral to the non-moving hand (ipsilateral to the moving hand). The ipsilateral activation was significantly higher in mirror conditions than the ipsilateral activation in direct conditions. Consistent with Experiment 1, Experiment 2, which was conducted with people with right upper limb amputation, showed higher ssMRCP amplitude in mirror view conditions compared to direct view conditions at the hemisphere ipsilateral to the actual (volitional) movement (contralateral to the reflected image of movement appearing as the hand on the amputated side). Ipsilateral activation in the mirror view condition was consistently higher than ipsilateral activation associated with the direct view condition at frontal (F3), central (C3) and parietal (P3) electrode sites. Experiment 3 revealed that visual observation of hand movements affects M1 excitability in healthy individuals. It showed that visual perception of own hand movements facilitates M1 activation whereas the performance of same hand movement with closed eyes inhibits M1 activation.
In summary, the present study provides clear evidence of M1 activation in association with a non-moving hand (which is visually perceived as moving). This unambiguous demonstration of M1 activation in association with the visually-perceived (non-moving) hand, suggests that perception of hand movement can directly lead to M1 activation. This thesis proposed a direct visual-to-motor activation mechanism to explain how visual feedback, particularly MVF, modulates sensorimotor activity. The direct visual-to-motor activation mechanism suggests that perception of hand movement reduces the activity of inhibitory networks in M1 that in turn facilitates M1 activation.