Computer-guided mental practice in neurorehabilitation.

Andrea Gaggioli, Emily A. Keshner, Patrice L. (Tamar) Weiss, Giuseppe Riva (Eds.) 195
Advanced Technologies in Rehabilitation.
Empowering Cognitive, Physical, Social and Communicative Skills through
Virtual Reality, Robots, Wearable Systems and Brain-Computer Interfaces
Amsterdam, IOS Press, 2009

© IOS Press, 2009


Computer-Guided Mental Practice in
Neurorehabilitation
Andrea GAGGIOLIa,b, Francesca MORGANTIb, Andrea MENEGHINIc, Ilaria
POZZATOd, Giovanni GREGGIOd, Maurizia PIGATTOc,d, Giuseppe RIVAa,b
aPsychology Faculty, Catholic University of Milan, Italy
bApplied Technology for Neuro-Psychology Lab, Istituto Auxologico Italiano, Italy
cAdvanced Technology in Rehabilitation Unit, Padua Teaching Hospital, Italy
dPadua University, Italy
Abstract. Motor imagery is the mental simulation of a movement without motor
output. In recent years, there has been growing interest towards the application of
motor imagery-based training, or “mental practice”, in stroke rehabilitation. We
have developed a virtual reality prototype (the VR Mirror) to support patients in
performing mental practice. The VR Mirror displays a three-dimensional
simulation of the movement to be imagined, using data acquired from the healthy
arm. We tested the system with nine post-stroke patients with chronic motor
impairment of the upper limb. After eight weeks of training with the VR Mirror,
remarkable improvement was noted in three cases, slight improvement in two
cases, and no improvement in four cases. All patients showed a good acceptance of
the procedure, suggesting that virtual reality technology can be successfully
integrated in mental practice interventions.
Keywords. Motor imagery, mental practice, stroke, rehabilitation, virtual reality
1. Introduction
Motor imagery refers to the mental simulation of a motor act in the absence of any
gross muscular activation [1]. The mental process of motor imagery has been
investigated within different areas of research, such as cognitive psychology,
neuroscience and sport psychology, sometimes with different terminology. In the
context of athletic performance studies, a frequently used concept is mental practice.
This term refers to a training technique by which a motor act is cognitively rehearsed
with the goal of improving performance. It is important to distinguish this specific
definition from the broader term mental preparation, which includes a variety of
disparate sport psychology techniques that share a goal of enhancing performance, such
as positive mental imagery, performance cues/concentration, relaxation/activation, self-
efficacy statements, and other forms of mental training. A distinction also needs to be
made between the “external” and “internal” perspectives in motor imagery. The
external perspective, considered to be mainly visual in nature, involves a third-person
view of the movement, as if watching oneself on a screen. The internal (or kinaesthetic)
perspective, on the other hand, requires a subject to take a first-person view and to
imagine the somesthetic feedback associated with action [2].
Recent studies in neuroscience have provided robust evidence that mental practice
with motor imagery may induce plastic changes in the motor system similar to actual

196

physical training [3, 4]. This supports the idea that mental training could be effective in
promoting motor recovery after damage to the central nervous system. In this chapter,
we first provide the rationale for using mental training in neurorehabilitation. Next, we
describe results of a pilot clinical trial, in which we examined the technical and clinical
feasibility of using virtual reality technology to support mental practice in stroke
recovery.
2. Motor imagery
Scientific investigation of motor imagery dates back to 1885, when the Viennese
psychologist, Stricker, collected the first empirical evidence that overt and covert motor
behaviours involve the same processing resources [5]. Over the past thirty-five years, a
number of studies have investigated this hypothesis further, by means of behavioural,
psycho-physiological and neuroimaging methodologies. Overall, these studies have
provided robust evidence about the existence of a striking functional similarity between
real and mentally imagined actions.
2.1. Chronometric studies
Chronometric studies are based on the Mental Chronometry paradigm, which involves
comparing real and imagined movement durations. In general, results of these studies
indicate a close temporal coupling between mentally imagined and executed movement.
Decety and Michel [6] compared actual and imagined movement times in a
graphic task. They found that the time taken by right-handed subjects to write a
sentence was the same whether the task was executed mentally or physically. Also,
subjects took approximately the same time, both physically and mentally, whether they
wrote the text in large letters or in small letters. This observation suggests that the
“isochronic principle”, which holds for physically performed drawing and writing tasks,
applies also to mentally-simulated motor tasks.
In another experiment, Decety and Jeannerod [7] investigated whether Fitt’s law
(which implies an inverse relationship between the accuracy of a movement and the
speed with which it can be performed), applies also to imagined movements. These
authors investigated mentally simulated motor behaviours within a virtual environment.
Participants were instructed to imagine themselves walking in a computer-
generated three-dimensional space toward gates of different apparent widths placed at
three different apparent distances. Results showed that response time increased for
decreasing gate widths when the gate was placed at different distances, as predicted by
Fitt’s law. According to authors, these findings support the hypothesis that mentally
simulated actions are governed by central motor rules.
The temporal correspondence between real and imagined motion is affected by
moderating variables such as the type of motor task and the time of the day. Rodriguez
and colleagues [8] asked a group of healthy subjects to perform or imagine a fast
sequence of finger movements of progressive complexity. Findings showed real-mental
congruency in relatively complex motor sequences (4 to 5 fingers), while in the
simplest sequences (performed with 1 to 2 fingers) real-mental congruency remarkably
decreased. The influence of the time of the day on real-mental congruency was
investigated by Gueugneau and colleagues [9]. They found that the real-virtual
isochrony was only observable between 2 pm and 8 pm, whereas in the morning and

197

later in the evening, the durations of mental movements were significantly longer than
the durations of real movements.
2.2. Psycho-physiological studies
Further evidence of the functional similarity between physical and imagined
movements is provided by studies that have measured patterns of autonomic response
during mental simulation of effortful motor actions. Decety and colleagues [10]
measured cardiac and ventilatory activity during actual and mental locomotion at
different speeds. Data analysis showed a strict correlation between heart and respiratory
rates and the degree of imagined effort. For example, the authors found that the amount
of vegetative arousal of a participant mentally running at 12 km/h was similar to that of
a subject physically walking at a speed of 5 km/h. In another study, Decety and
colleagues [11] analysed heart rate, respiration rate and muscular metabolism during
both actual and mental leg exercise. During motor imagery, vegetative activation was
found to be greater than expected from metabolic demands. The authors explained the
additional autonomic activation as the involvement of central mechanisms dedicated to
motor control, which anticipate the need for energetic mobilization required by the
planned movement.
Bonnet et al. [12] investigated changes in the excitability of spinal reflex pathways
during mental simulation and actual motor performance. In their experiment, subjects
were instructed either to exert or to mentally simulate a strong or a weak pressure on a
pedal with the left or the right foot. Modifications in the H- and T reflexes were
measured on both legs by electromyography (EMG). Findings showed that spinal
reflex excitability during motor imagery was only slightly weaker than in the reflex
facilitation associated with the actual performance. A further interesting result of this
study was that the lateralization and intensity of the imagined movement significantly
modulated the EMG activity during motor imagery.
2.3. Brain imaging studies
A large body of recent research has investigated neural substrates underlying motor
imagery by comparing the brain activation that occurs during mental and physical
execution of movements. Taken together, results derived from these studies suggest
that imagining a motor act is a cognitive task that engages a complex distributed neural
circuit, which includes the activation of primary motor cortex (M1), supplementary
motor area, dorsal and ventral lateral pre-motor cortices, superior and inferior parietal
lobule, pre-frontal areas, inferior frontal gyrus, superior temporal gyrus, primary
sensory cortex, secondary sensory area, insular cortex, anterior cingulate cortex, basal
ganglia and cerebellum [13, 15].
The pattern of cerebral activation associated with motor imagery can be influenced
by the level of motor expertise. Ross and colleagues [16] used fMRI to evaluate motor
imagery of the golf swing of golf players with different handicap. Results showed
activation of cerebellum, vermis, supplementary motor area, as well as motor and
parietal cortices. Moreover, the authors found a correlation between increased handicap
of participants and an increased number of activated brain areas. According to the
authors of this study, increased brain activity may reflect a failure to learn and become
highly automatic, or be related to a loss of automaticity with the need for compensatory
processing.

198

A controversial point is whether different types of movement imagery (e.g., visual
and kinesthetic) involve distinct neural networks. By using EEG, Davidson and
Schwartz [17] observed different patterns of occipital and sensory motor alpha activity
during kinesthetic versus visual imaging. In particular, visual imaging was associated
with greater relative occipital activation. In a fMRI experiment, Guillot and colleagues
[18] found that visual imagery was correlated with activation of the occipital regions
and the superior parietal lobules, whereas kinaesthetic imagery yielded more activity in
motor-associated structures and the inferior parietal lobule. These results suggest that,
like physical motion, these two imagery modalities are mediated by separate brain
systems.
2.4. Clinical neuro-psychology studies
Further evidence in support of the functional equivalence hypothesis comes from
clinical neuropsychological studies, showing that motor imagery is not dependent on
the ability to execute a movement but rather on central processing mechanisms.
Impaired motor imagery was observed in patients with lesions in the parietal cortex
[19] and in patients suffering from Parkinson’s disease, which affects supplementary
motor area, prefrontal cortex and basal ganglia [20, 22]. In those patients, movement
velocity during both motor execution and motor imagery is slower compared to healthy
controls; in contrast, patients with spinal lesions only show longer of motor execution
times but the same duration of MI motor imagery [23].Reduced functional motor
imagery was also identified in stroke patients with contralateral and premotor lesions,
with particular reference to upper limb pointing and rotation tasks [24, 25].
Furthermore, it appears that both imagery accuracy and temporal coupling can be
disrupted after a stroke, a phenomenon that has been defined by Sharma and colleagues
[26] as "chaotic motor imagery".
3. Mental practice
In the previous section, we have reviewed evidence suggesting that the execution of
mental and physical actions obey the same biomechanical constraints and share similar
neuromuscular mechanisms. Another stream of research has investigated the effects of
mental rehearsal on motor skill learning. Laboratory experiments involving healthy
individuals have shown that motor learning can occur through mental practice alone,
and that the combination of physical and mental rehearsal can lead to superior
performance compared to physical practice only [27]. Positive effects of mental
practice have been reported in a variety of motor tasks and for different outcome
variables, including performance accuracy, movement speed and muscular force [28,
31].
Neuro-physiological studies have consistently shown that prolonged mental
practice induce plastic changes in the brain which are similar to those resulting from
physical training. Pascual-Leone and colleagues [3] used transcranial magnetic
stimulation to examine patterns of functional reorganization of the brain after mental or
physical training of a motor skill. Participants practiced a one-handed piano exercise
over a period of five days. Results showed that the size of the contra-lateral cortical
output map for the long finger flexor and extensor muscles increased progressively
each day, and that the increase was equivalent in both physical and mental training.

199

Furthermore, both conditions produced performance improvements, although subjects
in the physical practice group displayed greater learning. However, the addition of one
physical training session allowed participants who practiced the task mentally to reach
the same level of performance as those who practiced physically.
Jackson and colleagues [4] used positron emission tomography to examine
functional changes associated with the learning of a sequence of foot movements
through intensive mental practice. The improvement of performance determined by
mental training was found to be associated with an increase in activity in the medial
aspect of the orbitofrontal cortex (OFC), and a decrease of activity in the cerebellum.
Data analysis also highlighted a positive correlation between the blood flow increase in
the OFC and the percentage of improvement on the foot sequence task.
Sacco and colleagues [32] used fMRI to measure the activity of brain areas
involved in locomotor imagery tasks (basic tango steps) at baseline and after one week
of training consisting of combined physical and mental practice. Findings showed an
expansion of active bilateral motor imagery areas during locomotor imagery after
training. Moreover, these authors found a decrease in visuospatial activation in the
posterior right brain, suggesting a decreased role of visual imagery processes in the
post-training period in favor of motor-kinesthetic ones.
3.1. Factors affecting mental practice
Other mental practice studies have examined the conditions under which this approach
is more effective. Driskell and colleagues [33] conducted a meta-analysis to determine
the effect of mental rehearsal and different moderators on performance. The key factors
highlighted by the review are summarized below:
- Type of task: mental practice seems to be more effective when the task to be
learnt require cognitive or symbolic components/operations (i.e. make decisions,
solve problems, generate hypotheses, p. 485);
- Retention interval: the effects of mental practice on performance become
weaker over time. To gain the maximum benefits of mental practice, one should
refresh training on at least a one- or two-week schedule (p. 489);
- Experience level: while experienced subjects benefit equally well from mental
practice, regardless of task type (cognitive or physical), novice subjects benefit
more from mental practice on cognitive tasks than on physical tasks (p. 488).
Mental practice may be more effective if novice subjects are given schematic
knowledge before mental practice of a physical task (p. 489);
- Duration of mental practice: the benefit of mental practice decreases with the
training duration. To maximize learning outcome, an overall training period of
approximately twenty minutes is recommended (p. 488).
The type of imagery modality (internal or external) used by the participant is
another important variable to consider when defining mental practice protocols. Fery
[34] found that in learning a new task, visual imagery is better for tasks that emphasize
form, while kinesthetic imagery is more suited for those tasks that emphasize timing or
fine coordination of the two hands. In another study, Hall and colleagues [35]
highlighted that kinesthetic imagery is better for learning closed motor skills, whereas
visual-based imagery is more effective for learning open motor skills.