Stroke neuroplasticity is most active in the first six months after a stroke. This window – the subacute phase – is when the brain is most open to reorganization. A new study from Gimcheon University tests a straightforward hypothesis: Does making rehabilitation tasks unpredictable push the brain to adapt more than predictable practice?
The researchers measured EEG activity and static balance in 30 subacute stroke patients across five weeks. The answer is clear, and it has direct implications for how clinicians approach stroke neuroplasticity in rehabilitation.
How the study was structured
Participants were randomly assigned to one of three groups. These were unpredictable task-oriented training (UG), predictable task-oriented training (PG), and conventional physical therapy (CG). Each group trained three times a week for five weeks, 30 minutes per session.
EEG was recorded using mbt’s mobile EEG device. Six electrode sites were monitored: two prefrontal (Fp1, Fp2), two frontal motor (F3, F4), and two parietal (P3, P4). These regions cover the circuitry most involved in executive function, motor planning, and sensorimotor coordination. They are also the regions most relevant to stroke neuroplasticity during motor rehabilitation.
The mobile form factor was essential here. Patients had gait disturbances and postural instability. EEG data for UG and PG was collected during a 10-metre walking task. And Smarting could easily move with the patient.
Balance was assessed on a BT-4 force platform. It measured sway area and sway length under eyes-open and eyes-closed conditions.
The same five tasks – completely different experience
Both trained groups completed the same five tasks. These were ball throwing and catching on a stable surface, the same on an unstable surface, 10-metre walking, ball catching during squats, and step box exercises. Each task ran for five minutes.
What differed was how each task was delivered. In the UG, balls arrived from unpredictable directions without warning. Mid-walk, the therapist called direction changes without notice. Step box height varied randomly – 15, 20, or 25 cm – between sets. The PG did the same tasks at fixed distances and constant speeds. Step height was a uniform 20 cm. Balls were announced before each throw.
The intervention table from the paper lays this contrast out clearly. Every row names the same task for both groups. But for UG, each task is defined by variability. For PG, it’s defined by fixed parameters. The CG received conventional physical therapy only: this included range of motion work, resistance training, and gait practice.
Alpha suppression: the brain staying switched on
Figure 1: Changes in relative alpha wave within and between groups. The changes are represented on all 6 monitored electrode locations.After five weeks, the UG showed the greatest alpha wave reductions across all six electrode sites. This is one of the clearest EEG markers of stroke neuroplasticity at work. Alpha suppression during task performance reflects active cortical engagement and information processing. When the next stimulus is unpredictable, the motor cortex can’t pre-program a response. It has to stay alert.
PG also showed alpha reductions, but smaller, while CG showed almost none. The between-group alpha data shows a consistent pattern. It runs across the full electrode array – from prefrontal to parietal. The effect isn’t localized to one region. It’s distributed across the cortex. That’s what you’d expect from training that demands whole-brain coordination.
Beta augmentation: the motor cortex working harder
Figure 2: Changes in relative beta wave within and between groups. The changes are represented on all 6 monitored electrode locations.Beta wave activity rose most in the UG, across all six sites. Beta augmentation reflects stronger sensorimotor cortical drive. It was most pronounced in the frontal motor and parietal regions. The UG’s beta gains were roughly double those of PG at most electrode sites.
PG showed moderate increases and CG showed minimal change. Together, the alpha and beta results point to enhanced cognitive-motor integration. This is a core mechanism of stroke neuroplasticity. The brain wasn’t just repeating movements, it was continuously adapting to meet demands it couldn’t anticipate.
Balance outcomes – and what the eyes-closed results reveal
Both trained groups improved their static balance, but UG improved more. Under eyes-open conditions, UG showed larger reductions in sway area and sway length than PG. CG showed no significant change.
The eyes-closed results are more striking. UG showed substantially larger sway reductions without vision than either comparison group. This matters clinically. Stroke patients typically over-rely on visual input for postural stability. Better performance with eyes closed points to a recalibration of sensory weighting. The brain is learning to use somatosensory and vestibular information; it’s no longer defaulting to vision. That’s a more robust and transferable form of stroke neuroplasticity for daily life.
The balance data tables break down sway area and sway length across both conditions. The gap between UG and the other groups is visible in both. It widens in the eyes-closed columns. That’s where the advantage of unpredictable training becomes hardest to ignore.
What this tells us about stroke neuroplasticity
Stroke neuroplasticity isn’t just about repetition volume. This study makes a direct case. Adding unpredictability to task-oriented training drives greater cortical adaptation and better balance recovery than predictable practice. The tasks themselves didn’t change – only their delivery did. Both EEG and balance data support this conclusion.
For clinicians, the implication is practical. Variability isn’t just variety for its own sake. It’s a neurological stimulus. Building it into rehabilitation during the subacute phase may be the most direct way to harness stroke neuroplasticity. The recovery window won’t stay open indefinitely.
You can find the original paper linked here!
Curious if this setup is right for your study?
Our team works with researchers across neuroscience, sports science and clinical applications - we'll help you figure out if mobile EEG fits your study design.
