Modulating the motor system by action observation: implications for stroke rehabilitation

Stroke is the leading cause of disability among adults (NINDS, 2006), and motor deficits of the arm and hand are a major contributor to functional disability following stroke (Sathian, et al., 2011). Stroke rehabilitation usually involves intensive motor practice to promote adaptive plasticity of structure and function in the motor system toward recovery (Nudo, 2006), to minimize motor deficits and develop new strategies in motor learning (Winstein & Wolf, 2008). In patients with poor motor ability, however, it is a challenge to provide relevant input for experience-dependent neuroplasticity. One way to drive plasticity in patients with limited mobility is suggested by the putative human mirror neuron system – cortical motor regions that respond when we perform an action and when we observe similar actions being performed by others (Rizzolatti & Craighero, 2004). By engaging motor circuits for observed actions in the mirror neuron system, methods in stroke rehabilitation, such as action observation, may help to rebuild motor function despite impairments, as an alternative or complement to motor practice. Recent theoretical (Buccino, Solodkin, & Small, 2006; Garrison, Winstein, & Aziz-Zadeh, 2010; Rizzolatti, Fabbri-Destro, & Cattaneo, 2009) and empirical evidence (Ertelt, et al., 2007; Franceschini, et al., 2010) supports the use of action observation after stroke. While this important beginning work shows that action observation has a positive impact after stroke, both basic and applied studies are limited, and lacking in evidence-based guidelines. The work presented here is the first set of studies to assess how stroke and stroke-related motor deficits affect cortical motor activity during action observation. ❧ The first study used functional magnetic resonance imaging (fMRI) to measure activity in motor-related brain regions during action observation. 12 participants with chronic middle cerebral artery stroke and moderate to severe dominant right hand paresis, and 12 matched right-handed non-disabled participants observed precision reach to grasp actions (e.g. lift pencil) made using the left and right hand. Observed actions were difficult or impossible for participants with stroke to perform using the paretic right hand, but easy to perform using the non-paretic left hand. All participants performed the actions using each hand to the best of their ability after the MRI. We find that non-disabled participants show bilateral, symmetric activation of cortical motor regions during left or right hand action observation. After stroke, a similar bilateral, symmetric activation of cortical motor regions is found during left hand action observation; yet during right (paretic) hand action observation, cortical motor activity is lateralized toward the left lesioned hemisphere. Overall, for non-disabled participants, left hand more than right hand action observation engaged cortical motor regions and more so in the right hemisphere; whereas for participants with stroke, right (paretic) hand more than left hand action observation engaged cortical motor regions, and more so in the left lesioned hemisphere. In addition, we find that activity in the motor system during action observation is related to motor capability to perform the observed actions, such that longer movement times using the paretic right hand, indicating great impairment, are associated with greater activity during right hand action observation in the inferior frontal gyrus of the left lesioned hemisphere. Results suggest that despite chronic non-use, cortical representations of the paretic limb in the damaged motor cortex are preserved and may be accessed by action observation in stroke rehabilitation. ❧ The second fMRI study assessed how activity in the putative mirror neuron system and other cortical motor regions during action observation differs between participants with stroke and different lesion locations. 6 participants with stroke involving the internal capsule, and 6 participants with stroke broadly involving the cortex and internal capsule, observed reach to grasp actions made using the left and right hand. All patients had chronic middle cerebral artery stroke of the dominant left hemisphere and moderate to severe right hand paresis. Results indicate a consistent finding related to stroke that is independent of lesion information in this study: participants with stroke show strong cortical motor activity in the left lesioned hemisphere during right (paretic) hand action observation. Yet within the overall stroke group pattern, we find differences between lesion groups, indicating a specific effect of lesion on MNS activity after stroke, including: 1) stroke involving the cortex and internal capsule is associated with more widespread, bilateral activity than stroke limited to the internal capsule; 2) stroke involving the left ventral inferior frontal gyrus (IFG) is associated with greater activity in the left pars triangularis of the IFG; and 3) for stroke involving the internal capsule, less motor capability to perform the observed actions is related to greater activity in the left IFG; whereas for stroke involving the cortex and internal capsule, this relation is found in the premotor cortex. Findings from this study suggest plasticity in the putative mirror neuron system to support action observation and imitation after stroke. ❧ The third study uses fMRI to evaluate whether action observation and execution share a common neural substrate after stroke affecting the motor system. 4 participants with chronic dominant left hemisphere stroke and moderate right hand paresis and 4 matched right-handed non-disabled participants took part in the study. During fMRI participants observed and performed a reach to grasp action (grasp tennis ball) using their left hand and right hand to the best of their ability. For each single subject we assessed activity during action observation, execution, and overlap between conditions. We find that single subject MNS maps are variable in the healthy brain and after stroke. We discuss contributions to variability related to age and stroke. We attempt to interpret single subject MNS maps related to stroke and related to the potential for a given individual with stroke to benefit from rehabilitative methods that engage the MNS. We find that MNS maps provide information that may be relevant to clinical applications, or may be used to evaluate cortical motor activity before and after an intervention that engages the MNS and related to functional gains. ❧ Overall, these studies provide first steps toward a neuroscience framework for the application of methods that engage the putative human mirror neuron system – such as action observation – for stroke rehabilitation.