1. Introduction

Stroke remains the leading cause of acquired adult disability worldwide, with motor impairment affecting approximately 80% of survivors in the acute phase [1]. While spontaneous neurological recovery is well documented within the first weeks post-onset, a large proportion of patients retain chronic deficits that profoundly impair quality of life and independence [2,3].

The neurobiological substrate of motor recovery involves plastic reorganisation of perilesional and contralesional cortical tissue, mediated by Hebbian strengthening of surviving neural circuits, axonal sprouting, and synaptogenesis [4]. Functional neuroimaging, particularly fMRI, has proven indispensable in mapping these changes longitudinally in living patients [5,6].

Despite a growing body of evidence, the precise trajectory of cortical reorganisation under structured sensorimotor training — as opposed to unguided spontaneous recovery — remains poorly characterised. The present study aimed to fill this gap by tracking whole-brain fMRI activation and resting-state connectivity in a longitudinal cohort of 84 patients undergoing a standardised 12-week training programme.

2. Methods

2.1 Participants

Eighty-four adults (47 male, 37 female; mean age 61.3 ± 9.2 years, range 42–78) with radiologically confirmed first-ever unilateral ischaemic stroke were recruited from three rehabilitation centres between January 2022 and June 2024. Inclusion criteria were: (i) stroke onset 2–8 weeks prior to enrolment; (ii) residual upper-limb motor deficit (Fugl-Meyer Assessment score 15–55); (iii) no contraindications to MRI. The study was approved by the institutional review boards of all participating centres (protocol ID: WUM/2021/0834) and written informed consent was obtained from all participants.

2.2 Sensorimotor Training Protocol

The 12-week programme comprised 45-minute daily sessions (5 days/week) combining task-oriented reaching tasks, proprioceptive perturbation exercises, and mirror therapy. Sessions were supervised by certified physiotherapists and were progressively intensified using a 3-phase periodisation model. Adherence was monitored via session logs and wearable accelerometry.

2.3 fMRI Acquisition & Analysis

MRI data were acquired on 3T Siemens Prisma scanners using a 32-channel head coil (TR = 2000 ms, TE = 30 ms, flip angle = 77°, 3 mm isotropic voxels). Resting-state and motor task (finger-tapping) paradigms were administered at baseline, week 6, and week 12. Preprocessing and statistical analysis followed SPM12 pipelines with permutation-based FWE correction at α = 0.05.

3. Results

At baseline, task-based fMRI demonstrated predominant contralesional M1 activation across the cohort, consistent with early compensatory recruitment. By week 6, a bilateral pattern was evident, with ipsilesional M1 reactivation in 61 of 84 participants (72.6%). At week 12, significant ipsilesional M1 activation was observed group-wide (z = 4.21, p < 0.001, FWE-corrected), accompanied by a reduction in contralesional signal (z = −2.87, p = 0.004).

Fugl-Meyer scores improved significantly over 12 weeks (baseline: 31.4 ± 10.2; week 12: 48.7 ± 9.8; t(83) = 14.6, p < 0.001). The change in ipsilesional M1 activation magnitude was strongly correlated with Fugl-Meyer score improvement (r = 0.61, 95% CI [0.47, 0.72], p < 0.001), indicating a neuroimaging biomarker of functional recovery.

4. Discussion

Our findings extend prior work by characterising the full 12-week trajectory of cortical reorganisation under a structured, high-dosage sensorimotor protocol. The progressive shift from contralesional to ipsilesional M1 dominance replicates cross-sectional observations [7,8] and aligns with the restorative-vs-compensatory framework of post-stroke plasticity proposed by Ward and colleagues [9].

Importantly, the correlation between cortical reorganisation and functional outcome (r = 0.61) suggests that ipsilesional M1 reactivation is not merely epiphenomenal but mechanistically linked to motor improvement. This observation has practical implications: fMRI-based biomarkers could potentially guide personalised rehabilitation scheduling and identify patients most likely to benefit from intensive training.

5. Conclusions

Twelve weeks of structured sensorimotor training drives a measurable, progressive shift toward ipsilesional M1 dominance in post-stroke patients, strongly correlated with functional motor recovery. These results support the development of neuroplasticity-informed rehabilitation protocols and highlight fMRI as a viable biomarker in clinical trial design.

References

1. GBD 2019 Stroke Collaborators. Global, regional, and national burden of stroke. Lancet Neurol. 2021;20:795–820.
2. Cramer SC. Repairing the human brain after stroke. Ann Neurol. 2008;63:272–287.
3. Langhorne P, Coupar F, Pollock A. Motor recovery after stroke. Lancet Neurol. 2009;8:741–754.
4. Nudo RJ. Neural bases of recovery after brain injury. Curr Opin Neurobiol. 2006;16:638–644.
5. Rossini PM, et al. Post-stroke plastic reorganisation in the adult brain. Lancet Neurol. 2003;2:493–502.
6. Stinear C. Prediction of motor recovery after stroke. Lancet Neurol. 2010;9:1228–1232.
7. Lotze M, et al. The role of multiple contralesional motor areas in post-stroke recovery. Brain. 2006;129:1474–1488.
8. Rehme AK, et al. Longitudinal changes in cortical activation during acute post-stroke recovery. Cereb Cortex. 2011;21:1383–1394.
9. Ward NS, et al. Neural correlates of motor recovery after stroke. Brain. 2003;126:2476–2496.