Page 87 - GTM-3-1

Global Translational Medicine BCI-FES with static magnetic field in SCI

of the training, classifier accuracy (AcCSP-LDA), and the higher the AcCSP-LDA, the greater the tendency to
number of misses and hits for the activation of the sFES, reduce the percentage of errors and increase the number
as shown in Figure 3. Some aspects that may have impaired of correct answers. It was possible to identify that in all
the learning effect of the training and the intervention sessions, errors were recorded in the activation of sFES,
with the sNES-sFES interface were to cancel out all the and in most cases, the percentage of errors exceeded the
factors that interfered with the participant’s concentration, percentage of correct answers. These errors occurred
such as maintaining absolute silence during training and in two situations: (i) activation of the sFES without
intervention, keeping them in a very comfortable position, the therapist’s command for this action, and (ii) non-
and avoiding any factors that increased muscle spasms. activation of the sFES when the motor imagery was
requested (within 4s). Regarding the activation of the
It was also observed that for the success of the training,
voice commands and orientations other than those mentioned sNES-sFES resulting from muscle spasms, it was not
counted as an error, as it possibly did not occur due to
above are necessary. We emphasize the importance of a mental response by the participant but rather due to
keeping the muscles intact, such as the relaxed upper limbs, interference with the sNES measurement.
especially when providing the command of motor imagery
to perform knee extension. In addition, participants tend to 4. Discussion
close their eyes in an attempt to concentrate, which is not
recommended due to the change in the sNES. In addition 4.1. Neuromuscular assessment
to the time, the amount of training per session was counted, The lack of improvement in neuromuscular function
as only one training was rarely performed owing to different can be explained by the complexity of the condition in
situations, such as an alteration of the sNES or a low value SCI, especially in the chronic phase, and the fact that
of the cross-correlation. As a result, between one and three a short-term intervention protocol was carried out. In
training sessions were performed because, despite the contrast to the present study, Donati et al. conducted a
32
participant denying mental fatigue, when they passed three study involving eight individuals with chronic SCI and
training sessions in the session, the values of the cross- paraplegics (seven complete and one incomplete) who
correlation were substantially reduced. underwent gait neurorehabilitation through a BCI in the
long term (12 months). The intervention was divided
3.3.3.1. Classifier accuracy into six stages, combining the integration of traditional
The classifier accuracy (AcCSP-LDA) resulting from physical rehabilitation and multiple brain-machine
training tended to increase gradually throughout the interface paradigms to restore locomotion. At the end
sessions (Figure 3), which can be explained by the of neurorehabilitation, all participants demonstrated
participant learning effect. The average AcCSP-LDA value neurological improvements in somatic sensitivity in
was 67% (a satisfactory AcCSP-LDA value was considered various dermatomes. In addition, they exhibited recovery
to be closer to 100%). of voluntary motor control in muscles below the level of SCI
(assessed by sEMG), resulting in a marked improvement in
3.3.3.2. Hits and misses their gait index. In other words, greater intensity, frequency,
Figure 3 depicts the percentage of hits, misses, and and long-term training are important for recovery.
successes in the sFES activation. It was observed that
4.2. Spasticity
In this study, we observed a reduction in spasticity
throughout the sessions, which is already well-established
in the literature regarding the efficacy of electrical
stimulation in spasticity reduction. As shown in the
systematic review, Bekhet et al. identified different
33
stimulation parameters that can reduce spasticity after
SCI, with the primary outcome being the assessment of
spasticity using MAS. These authors identified that the
stimulation parameters were a frequency of 20 – 30 Hz, a
pulse duration of 300 – 350 μs, and a current amplitude
>100 mA. The results of this review were similar to those
of the present study, showing that sFESs reduced spasticity
Figure 3. User’s learning trend. by 45 – 60%, in addition to reducing EMG activity and
Note: Dashed line = Classifier accuracy. increasing range of motion.

Volume 3 Issue 1 (2024) 7 https://doi.org/10.36922/gtm.2285