Fine
tactile sensation plays an important role in motor relearning after
stroke, and participates not only in initiating effective motor
behaviors but also in fine-tuning subsequent movements for fine motor
control [1,2,3].
It is notable however, that there is a relatively marginal amount of
knowledge regarding its dynamics in the process of post-stroke
rehabilitation. This is mainly due to a lack of effective evaluation on
neural responses to fine tactile stimulation. On the one hand, the
traditional measures of fine tactile impairments in clinical practice
are disadvantageous in terms of reliability and repeatability without
direct cortical detection [3].
For example, the two-point discrimination test depends not only on the
pressure applied to the finger by the examiner to induce tactile
stimulation, but also on the cognitive and discriminative levels of
patients in terms of subtle differences due to the inherently subjective
nature of tactile sensation [3].
On the other hand, functional neuroplasticity widely occurs in multiple
brain regions, including local and remote areas with respect to the
lesion site reorganized after a stroke. This would further result in the
cortical reorganization and connectivity disturbance, as previously
reported in studies on motor functions [4, 5].
A redistributed pattern from the ipsilesional hemisphere to the
contralesional hemisphere is commonly observed during motor or cognitive
tasks in stroke participants [6, 7].
However, compared to the extensively studied motor impairments, little
is known about the neuroplasticity associated with sensory impairments
post-stroke. This is principally due to a lack of evidence regarding the
strategies of cortical recruitment particularly in the area of fine
tactile sensation.
There are some studies on resting-state
functional magnetic resonance imaging (rsfMRI) which examined the
changes of cortical recruitment in relation to the tactile impairment
post-stroke, as revealed by functional connectivity (FC) [8, 9].
For example, Bannister et al. exploited rsfMRI to examine the
relationship between the recovery of tactile sensation and the
resting-state FC following a stroke. The results indicated that the
changes of resting-state FC between somatosensory regions and
distributed regions, including vision and attention networks, were
associated with improved tactile sensation within the first 6 months
post-stroke [8].
Goodin et al. also used rsfMRI to investigate the effect of different
lesion sites in the hemispheres on the functional connectivity of
tactile sensation in stroke participants [9].
It was found that the patients with lesions in the right hemisphere had
greater intra-hemispheric connectivity from the ipsilesional primary
somatosensory cortex (S1) to inferior parietal regions than those with
left lesions and unimpaired controls. However, these studies revealed
only the alterations of static cortical networks during the
resting-state after stroke. Furthermore, fMRI is limited in terms of
temporal resolution, despite the advantages of higher spatial resolution
and deeper imaging of brain activities beyond the cortical level than
electroencephalogram (EEG). In this sense, the fMRI is inadequate when
seeking the detection of the cortical activities in transient tactile
stimulation, since the sensory neurons change their levels of
sensitivity to a constant stimulus over time, i.e., sensory adaptation [10].
Thus, the available results on tactile impairments post-stroke might
not be suitable to reveal the strategies of the alteration in cortical
connectivity during the tactile sensation, which is a typically
transient process [11].
In
comparison to fMRI, when evaluating the cortical connectivity, EEG
offers a higher degree of temporal resolution when seeking to capture
neural activities during transient tasks [12, 13]. In this regard, the EEG-derived FC [14, 15],
demonstrating the interaction of information among cortical regions,
has been proven to be effective in capturing the alteration of cortical
connectivity in transient motor tasks in stroke survivors [16, 17].
For instance, Strens et al. compared the EEG-derived FC during a 25%
maximal handgrip task in chronic stroke participants and unimpaired
persons [16].
The results revealed greater FC between the ipsilesional supplementary
motor area (SMA) and sensorimotor area in the stroke than the unimpaired
controls, which might have a dynamically compensatory effect for brain
lesion after a stroke. The EEG-derived FC has also been applied to
measure the post-stroke alteration in cortical connectivity during the
repeated finger extensions with a frequency of 1 Hz. It was found that
the intensity of FC between contralesional motor/premotor cortex and SMA
was increased in stroke subjects compared with the unimpaired controls [17].
Despite the successful evaluations of the EEG-derived FC for motor
neuroplasticity following stroke, its investigation on sensory
neuroplasticity has not been well carried out. Such an investigation
would have the potential to further develop current understandings of
the alteration in cortical connectivity in relation to the tactile
impairments post-stroke.
The alteration of cortical connectivity
in its functional structure can be visualized by the graph theory-based
approach from a network perspective [18, 19],
where the EEG channels at different cortical locations and their FCs
are topographically represented as nodes and links among them [20].
The graph theory analysis has been adopted to reveal the stroke-induced
changes in functional brain networks from local (e.g., single-node
connectivity) to global level (e.g., connectivity of the entire brain)
represented by indices at difference scales [21].
Thus the examination of the dynamic information processing and neural
communication during motor or cognitive tasks was facilitated [22].
De Vico Fallani et al. also examined the functional brain organization
in stroke subjects whilst engaging in the finger tapping, where
inefficient brain networks were found in stroke participants with a
lower capacity to integrate the information from remote brain regions
and a lower capacity of processing information in local brain regions,
compared with unimpaired persons [21]. Additionally, in a study by Philips et al. on persons with chronic stroke [23],
the reduction of graph theoretical indices represented by the
parameters of global efficiency, local efficiency and the density of
intrahemispheric FC on the unaffected hemisphere was found to correlate
with post-stroke motor improvements measured by the increments in the
upper-extremity portion of the Fugl-Meyer Assessment (FMUE) after a
physical treatment for 12 weeks. However, little has been done using the
graph theoretical analysis to understand the functional structure in
relation to the connectivity alteration in the brain following the
post-stroke tactile impairments.
The purpose of this study was to
investigate the post-stroke alteration of cortical connectivity in
response to fine tactile stimulation via the textile fabric by
EEG-derived functional connectivity analysis. Whole brain EEG was
recorded from 64 scalp channels in 8 persons with chronic stroke and 8
age-matched unimpaired controls before and during the unilateral forearm
skin contact with cotton fabric. Functional connectivity was then
estimated using the EEG coherence method [24].
The fabric stimulation induced functional connectivity (SFC) was
analyzed by means of a cluster-based permutation test based on the
estimated FC [25].
Furthermore, the multiscale topological properties of functional brain
networks were assessed using the graph theory-based method to reveal the
functional structure of the connectivity alteration in the brain during
transient fine tactile sensation after stroke on multiple levels.
Finally, the alteration of brain connectivity in relation to the tactile
impairments post-stroke was discussed in detail.
Alison M. Aries
Valerie M. Pomeroy
Julius Sim
Susan Read4 and 
Susan M. Hunter