Underlying neural mechanisms of mirror therapy: Implications for motor rehabilitation in stroke
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.173622
Source of Support: None, Conflict of Interest: None
Mirror therapy (MT) is a valuable method for enhancing motor recovery in poststroke hemiparesis. The technique utilizes the mirror-illusion created by the movement of sound limb that is perceived as the paretic limb. MT is a simple and economical technique than can stimulate the brain noninvasively. The intervention unquestionably has neural foundation. But the underlying neural mechanisms inducing motor recovery are still unclear. In this review, the neural-modulation due to MT has been explored. Multiple areas of the brain such as the occipital lobe, dorsal frontal area and corpus callosum are involved during the simple MT regime. Bilateral premotor cortex, primary motor cortex, primary somatosensory cortex, and cerebellum also get reorganized to enhance the function of the damaged brain. The motor areas of the lesioned hemisphere receive visuo-motor processing information through the parieto-occipital lobe. The damaged motor cortex responds variably to the MT and may augment true motor recovery. Mirror neurons may also play a possible role in the cortico-stimulatory mechanisms occurring due to the MT.
Keywords: Brain stimulation; cerebrovasular accident; cortical reorganization; motor control neuroplasticity
The science of mirror therapy (MT) is getting due attention in the management of half-sided paresis due to stroke. The technique was first introduced by Ramachandran and Roger-Ramachandran to manage phantom sensations among the subjects with a unilateral amputation.
The characteristic property of a mirror that reflects an image after altering its right-left orientation is a commonly known fact. In humans, the movement of right side of the body is primarily controlled by the left brain, and vice versa. The mirror feature and the brain behaviour collectively form the concept of MT. The paradigm allows an individual to perform the movements of an unaffected body part in front of the arranged mirror-box while hiding the affected part. The technique induces a visual illusion that appears to mimic the movement of the paretic part [Figure 1]. The perception, more than being a simple feedback mechanism, has been found to enhance motor recovery of the impaired body side.,, The neural mechanisms leading to the favourable improvement are components of a complex phenomenon that need to be explored and comprehended.
In healthy individuals, MT provokes a distinctive brain activation that is not possible otherwise. The brain hemisphere, contralateral to the moving limb, exhibits activation in the primary motor cortex and also an alteration in the inter- or intra-hemispheric inhibition. MT also induces muscle activity within the inactive affected limb. Among poststroke hemiparetic subjects, the mirror intervention induces balance within the ipsilesional primary motor cortex by increasing interhemispheric communication, normalizes asymmetrical electrical activities and activates specific areas such as the precuneus and the posterior cingulate cortex.,,,
The primary motor cortex, located in the frontal lobe, is one of the important areas of the brain responsible for execution of the movements contralateral to the body. The posterior parietal cortex (movement planning), premotor cortex (movement observation) and supplementary motor area (bimanual coordination) also coordinate to control the movement. The cerebellum plays a key role in motor control and motor learning. The neural tracts descending from these areas get nearer in the white matter (as well as the internal capsule) and majority of the fibres cross at the brain stem level for further innervation to the contralateral body-side via the spinal cord while rest continue for innervation of the ipsilateral side.,,, Further, human brain is characterized by a complex neural network that establishes communication between multiple sensori-motor areas.
In stroke, interruption of the blood supply to specific brain areas leads to neuronal damage and consequently, an inability to execute voluntary movements (paresis) on the contralateral side. Due to the complex neural-networking, a small lesion may hamper the function of other non-lesioned areas. Further, due to uncrossing of a small proportion of the tracts, the ipsilateral body-side also exhibits a subtle motor impairment.,
It is now a well-established fact that the lesioned-brain has the ability to reorganize itself both structurally and functionally in response to certain sensory-motor experiences. The reorganization, referred to as neuroplasticity, may occur at various levels in the form of altered synthesis of brain-derived neurotrophic factors, synaptogenesis, cortical excitation and representation, and motor behaviour. Investigations using morphometry have also provided evidence of change in the amount of grey matter. The laterality index or asymmetry index (computation of the hemispheric dominance) using functional magnetic resonance imaging (fMRI) also confirms the existence of brain-plasticity induced by the motor therapies.
Various non-invasive therapeutic techniques have been developed to induce motor recovery based on neuroplasticity. Interventions such as MT, constraint-induced movement therapy, motor imagery, robotic training, and virtual training have different underlying neural-mechanisms. In addition to this, brain stimulation techniques such as repetitive transcranial magnetic stimulation and transcranial direct stimulation modulate cortical excitation associated with the positive motor recovery. Most of these techniques have been established based on their specific underlying neural mechanisms. Cost-effective methods such as MT, initially developed for the amelioration of phantom sensations in amputees, may also cause neural adaptations leading to a favourable motor control. The role of MT in facilitating true motor recovery rather than learned motor behaviour is justifiable. It is the only motor technique in which the affected limb is in resting position. Thus, the motor recovery of the impaired limb is definitely the result of neural reorganization at a certain level. However, its modus operandi for the induction of neuroplastic changes at different levels among the poststroke hemiparetic patients needs to be further explored.
MT has the potential to induce motor recovery in the paretic limbs of poststroke patients. In addition to the motor control, MT also leads to improvement in the activities of daily living. Many studies have investigated the effect of MT especially for the upper limb.,,,,, The technique has usually been made available in the form of specific movements, and is administered along with the conventional motor therapy. MT along with the standard motor therapy has been found to enhance the recovery and function of the upper limbs, especially the hand movements. Yavuzer et al., (2008) in a randomized trial of 40-subacute stroke subjects, found better (P < 0.01) recovery following the application of MT on Brunnstrom recovery stages (arm and hand) and Functional independence measure (self-care). In their systematic review, Thieme et al., (2013) concluded that MT was more beneficial in inducing motor functions (P = 0.002) and activities of daily living (P = 0.02) in comparison to the other forms of management. Further, the favourable effects of MT were present even after 6 months of the intervention. Samuel Kamlesh Kumar et al., (2014) studied the role of MT in improving upper limb function among chronic poststroke (>6 months) patients. The findings exhibited positive effect (P = 0.003 to 0.02) on the Fugl-Meyer assessment, the Brunnstrom recovery stages and the Box-and-block test. Some upper limb studies suggested that objects be utilized to impart MT., In a recent pilot trial on 33 poststroke chronic subjects, task-based MT along with standard motor therapy resulted in positive recovery (P < 0.001) for the paretic arm in comparison to the conventional method alone. MT has also been combined with other techniques such as afferent stimulation and transcranial direct stimulation to further augment the recovery.,, Lin et al., (2014) tested the somatosensory stimulation combined with MT in a group of 16 chronic stroke subjects. The therapeutic amalgamation resulted in favourable hand functions in terms of grasp and manual dexterity (P = 0.01 to 0.02). MT regime has also been applied to the paretic lower limb in a similar way using a mirror-frame placed between the legs in sitting position. In a lower extremity trial of 40 subacute-poststroke subjects, significant achievements were observed in the motor functions (P = 0.001 to 0.002). The experimental group received a 30-minute MT in the form of ankle dorsi-flexion movements of the sound side in addition to the conventional management. Post therapy, the group showed a favourable change in the lower-limb recovery and functional independence measures as compared to the controls. However, no significant difference was noted in the domains of spasticity and functional ambulation. In another MT trial on 34 subacute stroke subjects, a considerable improvement (P < 0.05; effect size 1.09 to 1.25) in gait was observed. The interventions comprised hip, knee and ankle movements of 20 sessions of 15 minute each along with the 30-minute standard rehabilitation. The temporospatial gait variables were assessed using a motion analyzer. The investigation found a positive change in the single-stance time, the step length and the stride length of the gait cycle in comparison to that of the controls.
The cortical activity due to MT has been adequately studied in the neurologically healthy individuals. The fMRI investigation revealed significant activation of the lateral occipital cortex (visual area), sensorimotor cortex, supplementary motor area and premotor cortex contralateral to the moving hand. In addition to this, the ipsilateral brain exhibited multiple loci of activation in the sensorimotor cortex, cerebellum and visual areas., Wang et al., (2013) investigated the brain activation using fMRI in response to MT (the right thumb and finger opposition movements perceived as the left sided one) among right-handed healthy individuals. The movements were recorded and displayed on a pair of special spectacles worn by the subject. The moving-hand was projected either as a mirror-image (right hand perceived as left) or horizontally inverted (software supported) normal image (right hand perceived as right). The findings exhibited lateralized activation of the right parieto-occipital cortex comprising the lingual gyrus, cuneus, middle occipital gyrus, and percuneus, in addition to bilateral motor areas. Further, on inverting the mirror-image (normal image) during MT, the activations shifted to the left-brain percuneus only. Thus, the perception of moving limb is the deciding factor for the cortical activity rather than the side of actual movement. The observation confirms the neural basis of mirror-illusion.
The excitability of ipsilateral primary motor cortex is also observable by numerous transcranial magnetic stimulation (tMS), magnetoencephalography (MEG) and motor evoked potential (MEP) findings.,,, The primary motor cortex gets robustly activated not only by moving a limb but also by viewing the mirror-image of the other limb as a perception of the first limb. In a MEG study, the 20-Hz rhythmic activity in the left Rolandic cortex (responsible for the right-hand motor performance) was analyzed in a group of healthy participants. The right median nerve was stimulated while holding a pencil or viewing its mirror image. The investigation revealed that the rhythmic cortical-activity got robustly restrained during the direct observation of right hand or mirror-induced perception of the left hand as the right side.
The ipsilateral cortex activation enhances with increase in the strength of illusion (for instance, keeping the moving-limb completely out of sight). A neumagnetic cortical investigation was conducted on 10 right-handed healthy volunteers while performing finger movements during MT. They perceived the mirror-image of the hand in 2 conditions, with and without viewing the moving hand. In the first condition, the activation of the primary motor cortex was recorded in 40% and 70% subjects respectively, for the dominant and non-dominant hand movements. However, in the second condition, all the participants exhibited cortical activity irrespective of the side.
Ipsilateral visual processing areas such as the anterior intraparietal sulcus (situated in the lateral parietal lobe) also demonstrated activation during MT. An fMRI investigation was conducted on healthy subjects during performance of finger and thumb movements. The right ipsilateral anterior intraparietal sulcus exhibited considerable number of activated voxels for right-sided movements while MT was being performed, when compared to the non-MT condition. The area controls the visuo-motor aspect of hand grasping and manipulation. It is considered to be responsible for further providing the information to reorganize the ipsilateral primary motor cortex. Other ipsilateral specialized visual areas such as the superior temporal gyrus and superior occipital gyrus also contribute to the processing of MT within the cortex. The superior occipital gyrus is linked to the posterior parietal cortex, which is also considered to be important for visuomotor functions. After multiple MT sessions, bilateral premotor areas increase functional connectivity with the contralateral primary sensorimotor cortex and supplementary motor areas.
In response to the dominant hand movements during MT, the non-dominant hemisphere experiences normalization of the asymmetrical activation. However, the movement of non-dominant hand leads to the recruitment of the mirror neuron system. Thus, MT enhances the contralateral (to the moving hand) brain to control movements in the resting limb.
The neural processing of MT may also be induced by intra- and inter-hemispheric inhibition. A tMS study was carried out to examine the training effect of MT (right hand activity perceived as the left sided one; 20 minutes, 4 days). Post-training, the intracortical inhibition reduced significantly in the right-ipsilateral primary motor cortex. Further, in comparison to the non-mirror condition, the left primary motor cortex demonstrated significant lowering of the intracortical inhibition after the mirror training. Additionally, a high gamma oscillation response (55-85 Hz) was also noted in the ipsilateral sensorimotor cortex of the moving hand in the mirror-condition. The observation was similar during the performace of real movement. Thus, MT can induce a feedback-control mechanism for movement perception even in the absence of actual proprioceptive inputs.
A recent electrophysiological study investigated the effect of MT-based wrist dorsiflexion movements using EEG. In comparison to the movements without mirror condition, the examination of low beta range frequency exhibited decrease in the event related de-synchronization in the brain contralateral to the moving hand, especially for the dominant side. In addition to this, the low mu range analysis revealed enhancement of the de-synchronization in both the brain hemispheres.
Bilateral activation of the primary somatosensory cortex also occurs during MT; however, the stimulation is more instant than that of the motor cortex. In response to the MT training, the neural pathway connecting the two brain-hemispheres (corpus callosum) was found to be excited. The communication amplified from the ipsilateral to the contralateral primary motor cortex, balancing the interhemispheric inhibition., The technique also enhances the excitatory role of corticospinal pathways corresponding to the non-moving limb.,
In addition to the neural activation of the ipsilateral brain, the muscles of the non-moving limb exhibit electromyographic contraction. The muscles correspond to that involved in performing the task by the other hand. The muscular activity of the passive limb increases with the performance time.
In poststroke hemiparesis, MT may augment the excitation of ipsilateral primary cortex associated with movements of the impaired limb. Further, the contralateral primary motor cortex may form new pathways for movement of the affected limb. Multiple MT sessions improve laterality index of the primary motor, premotor and supplementary motor areas of the ipsilesional hemisphere., In a phase II trial on 40 chronic hemiparetic subjects, the experimental group after a 6-week MT, demonstrated a change in activation pattern in the affected primary motor cortex. Specifically, precuneus and the posterior cingulate cortex (responsible for the self-control of movement and spatial relation) exhibited significant activation during MT comprising of bimanual movement (unaffected and affected-hidden limbs). The ipsilateral or contralateral cerebellum also gets activated. The ipsilesional hemisphere may become more active than the healthy brain., Bhasin et al., (2012) investigated the fMRI changes in response to computer-based MT (8 weeks) on 20 chronic stroke patients. Post therapy, considerable change (P < 0.05) in the laterality index of the ipsilesional primary motor and premotor cortex was observed among all the patients. In addition to this, increase in cluster activation was noted in the ipsilesional or contralesional cerebellum.
The movement-related beta desynchronization (measured by MEG) in the primary motor cortex is usually smaller among the poststroke subjects in comparison to the healthy individuals. However, within the hemiparetic subjects, the desynchronization is greater in the contralesional hemisphere than the ipsilesional cortex. Such asymmetry may be normalized by the MT. After the MT sessions, there is a decline in the activation of the contralesional brain and hemispheric activation balance moves toward the lesioned-brain. Thus, MT induces significant activation of areas exterior to the ipsilesional sensorimotor cortex. A functional interconnection, essential for the paretic limb movement, is also established between the ipsilesional somatosensory and motor cortex. In addition to the motor areas, MT also stimulates areas such as the dorsolateral prefrontal cortex related to cognition. Thus, cognitive involvement may also interact with the motor behavioural changes that are produced by the MT.
MT has a key component of movement observation. Mirror neurons, a set of cells in the premotor cortex and inferior parietal lobule get stimulated during observation as well as performing a motor act. The neural cells are responsible for the goal directed organization of movements. These systems of neurons help in reorganizing the damaged-brain and in enhancing motor control. Observation of the movements through MT may activate mirror neurons that further induce the closely-associated affected motor-cortex. Additionally, these neurons exhibit plasticity among healthy individuals. Thus, in hemiparetic subjects, observing movements through MT may exploit the mirror neuron system. The activation of the system is possible even during the acute phase or during the severe paretic stage of recovery when no voluntary movement exists. However, an in-depth study to establish the exact role and mechanism of mirror neurons in MT is warranted.
On the basis of various studies ,,,,,,,,,,,,,,,,,,,,, reviewed in this article, a comprehensive neuromechanism model for MT is proposed in [Figure 2]. The illustration exhibits MT for a left-hemiparetic (right brain damage) subject moving the right-unaffected upper limb and perceiving it as the left side. The model summarizes the involved brain areas and activities in response to MT. The visual perception of movement activates the occipital area on the lesioned side. The area may further project visuo-motor information either directly to the anterior intraparietal sulcus or from the superior occipital gyrus to the posterior parietal cortex. The dorsal frontal area participates to provide cognitive information related to the movement and its perception. After receiving visuo-motor messages, the primary somatosensory cortex on the lesioned side excites immediately while the premotor cortex activates later following many MT sessions, and the primary motor cortex responds variably. Multiple MT sessions reorganize bilateral premotor areas to establish functional communication with the contralateral primary motor cortex. An increased callosal communication from the damaged to undamaged brain also occurs balancing the interhemispheric inhibition. Bilateral cerebellum may also demonstrate activation to enhance motor control and motor learning. An increased excitability of corticospinal pathways corresponding to the affected body side is also observed. In addition to this, the primary motor area, primary somatosensory cortex, supplementary motor area, and premotor area of the non-lesioned brain shows stimulation in response to the movements performed in MT.
MT has a strong neural basis to be utilized in the motor rehabilitation of stroke. This simple and economical technique stimulates the brain noninvasively. In response to MT, numerous brain areas are harnessed to induce favourable neuroplasticity and associated motor recovery. The regime activates not only the areas corresponding to the moving hand but also the ipsilateral (to the moving hand) brain by means of the mirror- illusion of movement. Brain stimulation is possible among both the healthy individuals and poststroke hemiparetic subjects. MT may reorganize the damaged brain and balances the activity of both the hemispheres.
The author acknowledges Mr. Ankit Mittal for extending his support in illustrating [Figure 1].
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2]