Motor Recovery
Motor impairment after stroke is the major cause of permanent disability. Recovery of hand motor function is crucial in order to perform activities of daily living, but is often variable and incomplete (Duncan et al., 1992). Indeed, stroke rehabilitation efficacy is limited (de Pedro-Cuesta et al., 1992; Duncan, 1997) with 30 to 60% of patients unable to use their more affected arms functionally after discharge (Kwakkel et al., 1999; Lai et al., 2002). Currently, neuroscience-based rehabilitation seeks to stimulate spontaneous functional motor recovery by capitalizing on the inherent potential of the brain for plastic reorganization after stroke (Chollet et al., 1991; Netz et al., 1997; Platz et al., 2000; Feydy et al., 2002; Cramer, 2004; Dobkin, 2004; Ward and Cohen, 2004; Gerloff et al., 2006; Nudo, 2006).
In this regard, evidence from animal studies encourage the parallelism between plasticity mechanisms in the developing nervous system and those taking place in adult brain after stroke (Murphy and Corbett, 2009). On the other hand, understanding the effect of rehabilitative practices on brain plasticity has the potential to provide a neural substrate to underpin rehabilitation and hence, in developing novel rehabilitation strategies (Liepert et al., 2000).
Rehabilitative interventions aimed at functional motor recovery in stroke patients are based mainly on active movement training such as constraint-induced therapy and/or passive mobilization (Liepert et al., 2000; Schaechter, 2004; Wolf et al., 2006). Recent clinical trials have provided new insights into the methods to assist motor recovery after stroke (Dobkin, 2008; Langhorne et al., 2009; Subramanian et al., 2010). A recurrent theme is that interventions emphasizing intense active repetitive task-oriented movements are of high value in this regard. To promote the effects of training and practice, biomedical engineers, neuroscientists, and clinicians have started an intense joint collaboration over the past 10 years. This technological approach holds a promise for enhancing traditional post-stroke recovery in different ways: exercise in virtual environments could provide feedback to aid skills learning (Jack et al., 2001; Holden et al., 2005; Merians et al., 2006); robotic assistive devices with sensory feedback for repetitive practice could provide therapy for a long periods of time, in a consistent and measurable manner (Takahashi et al., 2008; Volpe et al., 2009); FES of muscles might enable movements not otherwise possible during the practice of tasks such as reaching to grasp an object (Alon et al., 2007). These are only a part of the increasing technological developments which have been recently applied in sample of stroke patients and showed the feasibility in providing a clear incremental reduction of motor impairments offering, therefore the opportunity to build a better outcome for patients.
These treatments are based on the ability of the patients to perform actions with the affected hand or arm and therefore, require residual motor ability. Many patients however, are prevented from training based on the above treatments due to having no residual hand motor functions. In case of moderate to severe motor deficits, MI represents an intriguing new “backdoor” approach to access the motor system and rehabilitation at all stages of stroke recovery (Sharma et al., 2006, 2009a,b; Page et al., 2007). MI can be defined as a dynamic state during which the representation of a specific motor action is internally rehearsed without any overt motor output, and that is governed by the principles of central and peripheral motor control (Decety and Jeannerod, 1995; Berthoz, 1996; Jeannerod and Frak, 1999; Lotze and Halsband, 2006). This is likely the reason why mental practice using MI training results in motor performance improvements (for a review in athletes, see Feltz and Landers, 1983; Dickstein and Deutsch, 2007). In addition, MI training can independently improve motor performance and produce similar cortical plastic changes (Lotze and Halsband, 2006), providing a useful alternative when physical training is not possible.
Despite this evidence, imagery training of movements combined with conventional physiotherapy of the hand has been reported in few structured clinical trials including subacute to chronic stroke patients and they demonstrated a greater improvement of hand function with the additional mental practice (Braun et al., 2006; Page et al., 2007; Malouin et al., 2008; Simmons et al., 2008; Verbunt et al., 2008). Up to now, no definite conclusions can be drawn, except that further research using a clear definition of mental practice content and standard outcome measurements are needed. As for the first point, it follows from the definition of MI that because of its concealed nature, a subject may surreptitiously use alternative cognitive strategies that, if not screened for, could confound investigations and produce conflicting results. Because the aim of MI is to activate the motor networks, it is crucial that subjects perform the mental task from the first person perspective (so called kinesthetic MI), in contrast to third person perspective or visual imagery (Decety and Grezes, 1999; Neuper et al., 2005). In this regard, a recent fMRI study on MI (Guillot et al., 2008) has looked at this issue by assessing subjects’ imagery abilities using well-established psychological, chronometric, and new physiological measures from the autonomic nervous system. The results suggest that visual and kinesthetic imagery are mediated through separate neural systems, which contribute differently during processes of motor learning and neurological rehabilitation.
Beyond these overall considerations, the challenge neurorehabilitators are faced with is clear: to modulate the sensorimotor experience of stroke patients to induce specific form of plasticity to boost relearning processes. Pulling all previous evidence together, a promising and challenging approach is to deploy BCI technology as a tool to tackle the challenge in the field of functional motor recovery after stroke. Indeed, the inherent BCI training paradigms will be exploited as a behavioral, controlled strategy to recruit and/or reinforce patient's sensorimotor experience (like MI and/or residual motor ability) during functional motor recovery after stroke and, thus to enhance those physiological plasticity phenomena which are the substrate for the functional motor recovery itself. The feasibility and effectiveness of a BCI-based neurofeedback paradigm will be enhanced by combining MI with motor action observation; this latter cognitive strategy will be allowed via technology such as visual representation and FES of the hand. Moreover, a multimodal brain imaging approach will provide detailed knowledge of how the brain encodes and processes information when it imagines the control or actually controls a peripheral device. This knowledge will, in turn, unravel to what extent long-term use of BCI “per se” affects the brain activity of the user.
The BCI community has a long-standing experience with one of the employed strategies for operating EEG-based BCI systems – the modulation of sensorimotor EEG reactivity induced by movement imagery tasks (Pfurtscheller et al., 1997; Neuper et al., 1999, 2006; Cincotti et al., 2003; Kübler et al., 2005). This makes possible the development of flexible and affordable BCI tools to objectify and to monitor individual MI execution both in terms of performance (relation between subject MI performance and subject level of accuracy in controlling BCI-operated basic applications) and compliance (identification of a correct MI task which is needed to achieve BCI-system control).
Within the BCI community, the opportunity to use BCI protocols to promote recovery of motor function by encouraging and guiding plasticity phenomena occurring after stroke (or more generally after brain injury) is at a very preliminary stage (for review see Birbaumer et al., 2008; Daly and Wolpaw, 2008; Mak and Wolpaw, 2010). Discussion is currently underway over several factors including: the extent to which patients have detectable brain signals that can support training strategies; which brain signal features are best suited for use in restoring motor functions and how these features can be used most effectively; and what the most effective formats are for the BCIs aimed at improving motor functions (for instance, what guidance should be provided to the user to maximize training that produces beneficial changes in brain signals). So far, preliminary findings are promising: Scherer et al. (2007a) suggested that event-related EEG activity time-frequency maps of event-related EEG activity and their classification are proper tools to monitor MI related brain activity in stroke patients and to contribute to quantify the effectiveness of MI. Buch et al. (2008) have shown that six out of eight chronic stroke patients suffering from a handplegia learned to control a magnetoencephalography-based BCI by MI. In all these cases, the best signals were depicted over the ipsilateral (unaffected hemisphere). Other attempts to use non-invasive BCI for rehabilitation include Ang et al. (2009) and Prasad et al. (2009). Finally, the idea that BCI technology can induce neuroplasticity has received remarkable support from the community based on invasive detection of brain electrical signals (for recent review see Wang et al., 2010).
As mentioned above, a general consensus from the clinical point of view is still lacking on the content, dose, and strategy of the MI intervention in stroke rehabilitation. What's more, there is no evidence so far, that one intervention protocol can be more effective with respect to another, for the mental practice of motor actions. According to the extensive review by Sharma et al. (2006) only few studies have paid attention to the previous issues and the conclusion can be summarized as follows: (i) MI training has to be provided in addition to a background rehabilitation therapy; (ii) MI tasks should be practiced in the patient's functional context to be most effective; in this regard, the MI tasks can be chosen from activities of daily life (i.e., reaching for and grasping a cup or other objects, turning page in a book, proper use of writing tool) from the content of the occupational therapy (Page et al., 2007). A more recent approach suggests MI interventions to be tailored on specific individual possibilities, skills, and needs of the patient in accordance with evidence-based practice (Braun et al., 2008).
Finally, the measurement of the impact of new rehabilitative interventions on patient motor impairment is another issue of utmost importance. One valuable instrument which can offer a solid way to generalize results obtained from clinical/research trials, is represented by International Classification of Functioning (ICF). In a recent study, the effectiveness of a mental practice-based training on post-stroke rehabilitation has been evaluated by considering primary and secondary outcome measures according to the ICF domains (impairment; activity; participation and quality of life) (Verbunt et al., 2008).