Review Article ISSN 2641-4317

International Journal of Psychiatry Research

Neurorehabilitation: Recovery Advances through CNS Neuromodulation

Denis Larrivee1,2*
*
Correspondence:
Loyola University, Chicago.
1 Denis Larrivee, Loyola University, Chicago and University of
Navarra Medical School, Pamplona, Spain.
University of Navarra Medical School, Pamplona, Spain.
2

Received: 29 November 2019; Accepted: 22 December 2019

Citation: Denis Larrivee. Neurorehabilitation: Recovery Advances through CNS Neuromodulation. Int J Psychiatr Res. 2019; 2(7): 1-4.

ABSTRACT
Modern neurorehabilitation promises to revolutionize standard interventions for CNS impairments in domains
traditionally dominated by physiotherapy. Functional neuroimaging models have greatly increased understanding
of underlying mechanisms contributing to brain pathology, enabling improved diagnosis and targeted therapy.
Building on diagnostic clarity, a spectrum of technological advances, ranging from machine learning and BCI
intervention to non-invasive neurostimulation, now assist functional recovery either directly, through modulation
of innate circuit and molecular plasticity, or indirectly, through externally controlled motor support.

Keywords activity signatures. In turn, improved diagnosis grounds targeted

Brain computer interfacing, Machine learning, Neurorehabilitation, intervention, that increasingly relies on machine learning methods
Neuromodulation, Transcranial direct current stimulation. to interpret semantic content and on neurostimulation protocols
to modulate molecular and circuit-based brain plasticity. Together
Introduction these advances promise significant hope for millions of individuals.
Among the most significant and actively investigated domains into
CNS impairments, neurorehabilitation promises to revolutionize Diagnostic Models
standard approaches typically dominated by physiotherapies. Modern diagnostic protocols extrapolate from a growing evidence
Submission requests to PubMed as of October 2019 [1] retrieve base showing that functionally related activity is distributed
entries totaling in excess of 54,000, of which 5000 have been over broad neural landscapes, often globally [7]. For CNS
generated per annum during the last five years, a fourfold increase impairments, experimental paradigms therefore seek to monitor
over some 1,200 per annum a decade ago, and fifty fold greater intrinsic activity relations between widely separated regions,
than the 100 articles per annum produced at the field's modern or attempt to detect differences arising from globally induced
reincarnation in 1995. activity, usually in response to task based paradigms. Such
paradigms vary from simple demonstrations of task correlated
Underlying factors are multiple but begin with the understanding activity to complex dependency relations, where activity in one
that brain dysfunctions are widely variable and highly prevalent zone is 'causally' linked to activity in one or multiple other brain
in the general population. Vascular disorders such as ischemic regions. The frequently used general linear model assumes, for
strokes or subdural hemotomas, degenerative diseases like example, that the observed activity changes are multifactorial and
Parkinson's (PD), Amyotrophic Lateral Sclerosis (ALS), or related to multiple independent variables, where recorded pixel
Multiple Sclerosis (MS), infections like meningitis and trauma, values from imaging data are equated with linear combinations
and structural and functional disorders all contribute to brain of explanatory variables [8-10]. Numerous such paradigms have
pathologies affecting motor performance [2,3]. Cumulatively, been developed. Among the best known and most frequently used
despite considerable variability in impairment type and incidence, are the Psycho Physiological Interaction (PPI), Dynamic Causal
underlying mechanisms for many such dysfunctions are being Model (DCM), Granger Causal Model, and Multivoxel Pattern
elucidated. This knowledge has considerably increased diagnostic analyses, which offer distinct advantages depending on research
power. Structural and physiological diagnoses, chiefly by means of and diagnostic objectives. The DCM, for example, estimates
sophisticated mathematical models of neuroimaging data [4-6], for effective connectivity and the influence on connectivity of tasking
instance, can identify different diseases on the basis of characteristic variables. Its approach involves the building of a reasonably
Int J Psychiatr Res, 2019 Volume 2 | Issue 7 | 1 of 4
realistic model of interacting brain regions and then assessing communicating with the brain, it does signify an advance over
how this model would be transformed by influencing variables existing imaging techniques in exposing the structure of the
under task based circumstances. In the Grainger causal model, by information content that the brain may actually be using. This
contrast, estimates of causal origin assume that such influences has direct relevance for assessing communication errors that may
exhibit temporal precedence. Data analyses therefore search for underlie cognitive impairments. The monitoring of low frequency
time shifted versions of activity patterning between different brain brain oscillations, for example, has recently been used to track
regions. Together these approaches provide considerably improved motor recovery following stroke [17].
insight into neurophysiological correlates of CNS impairments.
In the ‘decoding’ approaches evolved to date the central technical
Therapeutic Options for Assisted Functional Recovery concern is that of ‘classification’, that is, mapping a brain state
The improvements in diagnostic capability and the significant imaged in its activity pattern with an external feature, object, or
need for intervention have been the stimulus for development event. The approach having the longest history, mass univariate
of the current spectrum of therapeutic approaches [11]. Often analysis, is based on a general linear model in which sequential
incorporating various neurotechnologies they directly target the brain regions are monitored for specific mental activity at a
CNS itself, modulating circuit-based connections at the level specific brain location [18]. While it is not known how the brain
of higher order, cognitive functions. In consequence, there is represents mental content, it is presumed that it is distributed over
a growing consensus that functional recovery is in many cases populations of cells [19]. Hence, there is the presupposition of an
achievable, either indirectly, through technically enhanced, assisted underlying connectivity architecture uniting them.
replacement of motor abilities, or directly, by neural restoration.
Classification technology is used to measure the covariance
Of these two routes to recovery, restoration of nerve tissue between multiple single units, which serves as a diagnostic feature
remains the gold standard. Nonetheless, in many cases this option that is relevant to how select images are encoded. In fMRI imaging,
is precluded by the permanence of cortical nerve tissue damage. for example, the presentation of a single object will activate long
Because of the inability of most CNS neurons to divide and regions of the occipital cortex originating at multiple sites; thus,
replicate due to their normally arrested cell cycles, functional brain monitoring covariance is thought to relate neural activity patterns
tissue is generally incapable of replacing lost neurons in damaged to a structured representational content [20].
areas. Bypassing damaged tissue thus often remains the sole option
for functional recovery, one made increasingly tractable through BCI Contacts
advances in understanding the computational language of the brain Notwithstanding the significance of interpretive paradigms,
[12]. coupling to external devices that can translate the meaning of
commands is essential for implementing assistive technology.
The brain's language is currently thought to be composed from Current methods for transferring this information rely on brain
cyclical activity that is generated by groups of neurons through computer interfacing (BCI), which has evolved considerably
feedback and feedforward neural circuits yielding temporally since Jacque Vidal first coined the term in the 1970s [21]. Recent
independent and patterned features [13]. These features are methods have been used for rehabilitation of stroke victims,
largely non-linear and dynamical and emerge from the high- improved learning with artificial sensory feedback, and real-time
dimensional state space that characterizes the global activity of control over fine motor movements among others.
the brain. Accordingly, they have the potential for generating an
indefinite number of syntactical elements that can be combined For neurorehabilitation of cognitive and CNS impairments, BCI
and recombined to construct arrays yielding various neural codes. is a theoretical outgrowth of several generations of endogenous
Simple features, like fixed point attractors [14], for example, are devices that have as a prime strategy the direct replacement of
mathematically described by linear relations between the rate of lost neural function. Among the many devices developed for
change of the attractor’s return to its original configuration and replacement of nerve function outside the brain include pacemakers,
the brain state, typically represented by a signal feature related to cochlear implants, and vagal stimulators, for instance, which have
that state. More complex models, which can be mathematically all been successfully deployed in the relatively simpler anatomical
described by multiple parameters [15], make the language substrate of sensorial and motor nerves [22]. Cochlear implants,
exceptionally complex. for instance, transduce pitch vibrations that occur outside the ear
to coded electrical signals within the cochlea in order to elicit
It is with the intention of interpreting the semantic content of action potentials in the frequency to place receptors that form the
brain activity from such neural codes, as opposed to making auditory nerve.
functional inferences about brain state activity, that qualitatively
new approaches for elucidating what brain states actually mean Among non-invasive BCIs sensory evoked potentials offer the
have been undertaken [16]. These new approaches attempt most direct channel mediating between the brain' computational
to reveal semantic content by correlating brain activity with language and the devices intended to carry out the brain's
objective features of the world. Although this is not quite the commands. Of these the steady state visual evoked potential
same as representational imagery of the sort needed for directly (SSVEP) is generally recognized as the most easily observed and
Int J Psychiatr Res, 2019 Volume 2 | Issue 7 | 2 of 4
most accurate representation of brain-based information [23]. The Conclusion
SSVEP is an EEG recorded signal phase locked to the subject's Although numerous mechanisms can impair CNS motor
attended visual stimulation. Accordingly, the accuracy of BCI performance, therapeutic methods offer increasingly tractable
signaling requires carefully constructed algorithms for segregating solutions for repairing damaged brain tissue and restoring motor
stimulus dependent responses. Increasingly, these capabilities function. As a result, neurorehabilitation is no longer viewed as
emerge from sophisticated machine learning, artificial intelligence a domain of physical therapy alone, but a promising avenue for
technologies. Whole brain analysis for high dimensional data directly treating or bypassing underlying brain tissue damage.
employs machine learning approaches to discover multivariate Improvements in the understanding of the brain's language, the
relationships in data acquired from neuroimaging analyses [24]. ability to directly induce function through plastic change, and
Machine learning has been used, for example, to differentiate sophisticated data gathering and information processing abilities
among population groups and to predict behavioral outcome. More that can decode and transmit the brains signals into motor
recently, it has been used to identify neural correlates that can be execution, offer today's patients realistic recovery options not
targeted for stage specific BCI intervention [25]. available two or even a decade ago.

Neurorehabilitation and Brain Tissue Restoration References

Beyond the extraordinary growth in sophistication and 1. Klein RG, Mannuzza S, Olazagasti MA, et al. Clinical
range of capabilities for assisted recovery, a select group of and functional outcome of childhood attention-deficit/
neurorehabilitation procedures has also been used to successfully hyperactivity disorder 33 years later. Arch Gen Psychiatry.
restore normally executed nerve and motor function. Implicitly or 2012; 69: 1295-1303.
explicitly a primary goal of this research attempts to access the 2. McCarthy LF. Behavior therapy for ADHD children: more
brain's innate plasticity. Methodologies therefore attempt to evoke carrot, less stick. ADDitude Summer. 2008: 12-15.
plastic changes either directly by non-invasive neurostimulation, 3. Chien IC, Lin CH, Chou YJ, et al. Prevalence, incidence, and
or indirectly through patient directed reconfiguration of functional stimulant use of attention-deficit hyperactivity disorder in
channels, known as neurofeedback. In either case an important Taiwan. 1996-2005: a national population-based study. Soc
presupposition is the distributed nature of functional activity. Such Psychiatry Psychiatr Epidemiol. 2012; 47: 1885-1890.
effective connectivity often extends beyond the point of lesion, 4. Pan JH. Only 1 in 20 kids with ADHD get proper treatment.
where it can be molded to regenerate lost functional associations. Press conference by Taiwanese Society of Child and
Adolescent Psychiatry (TSCAP). Taipei: China Post. 2011; 1:
In the case of neurofeedback approaches, fMRI imagery is often 15.
used to monitor brain activity during therapeutic or training 5. Pontifex MB, Saliba BJ, Raine LB, et al. Exercise improves
paradigms [26], to assess the effectiveness of self-guided behavioral, neurocognitive, and scholastic performance
modulation. For motor tasks this requires the identification in children with attention-deficit/hyperactivity disorder. J
of functionally relevant activity distributed over several brain Pediatr. 2013; 162: 543-551.
domains, including motor cortex, premotor cortex, supplementary 6. Arnsten AF. Toward a new understanding of attention-deficit
motor area, parietal cortex, basal ganglia, and cerebellum. For a hyperactivity disorder pathophysiology: an important role for
patient suffering from ischemic stroke of the middle cerebral artery, prefrontal cortex dysfunction. CNS Drugs. 2009; 23: 33-41.
for example, identification of functionally overlapping regions 7. Gilliam JE. Attention-deficit/hyperactivity disorder test
allows the patient the prospect of viewing self-enhanced activity in (ADHDT). Psychological Assessment Resources. 1995; 25:
a region of interest and then monitoring the restoration of function 3-165.
in these regions over time. Current evidence suggests that using 8. Jung SS, Lee YC. Manual of child sensory-motor integration
even a single target region involved in motor imagery can lead to (SMI) checklist and norm. In Taipei Special Education Center.
changes in cortical and subcortical network connectivity. Elementary Special Education: Booklet. 1991; 32: 7-36.
9. Brown L, Sherbenou RJ, Johnsen SK. The Test of Nonverbal
Neurostimulation methods like transcranial direct current Intelligence, Second Edition. Austin Texas: PRO-ED
stimulation (tDCS), on the other hand, propose to modify brain Inc.1990; 1-55.
plasticity by directly stimulating nerve tissue [27]. Such approaches 10. American Psychiatric Association. Diagnostic and statistical
allow top down modification of cortical areas to restore motor manual of mental disorders (fourth ed. text rev). Washington
abilities through neuroplastic changes. Coupled with sophisticated DC: 2000; 83-85.
imaging procedures, the progress of restoration is related to the 11. National Institute of Mental Health. Attention deficit
appearance of neural correlates that relate progress to a succession hyperactivity disorder ADHD: A detailed booklet describing
of activity events. A burgeoning medical establishment is now ADHD symptoms, causes, and treatments, with information
devoted to non-invasive stimulation strategies, which have been on getting help and coping. U.S. Department of Health and
successfully used for pain. On the other hand, optimization of Human Services National Institutes of Health. National
parameters for motor recovery remains selective and overall Institutes of Health Publication No. 08-3572. 2008.
standardizations have yet to be achieved. 12. Ayres J. Sensory integration and the child. Los Angeles:
Western Psychological Services. 1980; 81-120.
Int J Psychiatr Res, 2019 Volume 2 | Issue 7 | 3 of 4
13. Jess M, Mahtob K, Ashley K. Body movement kinesthetic and Taiwan.
vestibular sense. Tangient LLC. 2016. 16. Jung SS, Jung S. A Sense of Integration Training Mode for 1~4
14. Jung SS, Yeh TZ, Loo HH, et al. Efficacy of fMRI in ADHD Years Small Children, with Very Sensitive, Crying a Lot, or
children with sensory-motor integration training. Taiwan Spec Under-Developments. How to Apply Morita Organs-Oriented
Ed Q. 2006; 101: 9-16. Therapy and Differential from sensory-Motor-Integration-
15. Jung SS, Jung S. Sensory-motor-integration training for Training. From J Pediatric Neonatal. 2019; 1: 1-5.
students with attention deficit/hyperactivity disorder in

© 2019 Denis Larrivee. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License

Int J Psychiatr Res, 2019 Volume 2 | Issue 7 | 4 of 4