Unveiling the upper-limb functional recovery mechanisms in stroke patients using brain-machine interfaces: a near-infrared functional imaging-based study

Upper limb dysfunction is highly prevalent among patients in the chronic stage of stroke. Brain-computer interface (BCI) technology, which creates a direct link between the brain's electrical signals and external devices, stroke patients with motor disabilities are able to perform BCI tasks for clinical rehabilitation. However, traditional BCI applications are often limited in their capacity to monitor the brain function of patients. In this study, functional near-infrared spectroscopy (fNIRS) was employed to observe changes in brain cortex activation patterns before and after BCI use in ischemic stroke patients with upper limb dysfunction. Thirty-four ischemic stroke patients with upper limb dysfunction meeting the inclusion criteria were selected and randomly assigned to either a treatment group or a control group using a random number table, with 17 patients in each group. During the study, 4 participants dropped out, leaving 30 patients for the final statistical analysis, 15 in each group. Both groups received routine upper limb rehabilitation training. Additionally, the treatment group underwent daily BCI training for 30 min, 5 days a week, for 4 consecutive weeks. Upper limb function was evaluated using the Fugl-Meyer assessment for upper extremity (FM), and daily living activities were assessed with the modified barthel index (MBI). The six regions of interest (ROIs) in the cortex for fNIRS measurement were the ipsilesional and contralesional primary motor cortex (PMC), supplementary motor area (SMA), and somatosensory motor cortex (SMC). The three time points of measurement were baseline (prior to any treatment), 2 weeks of treatment, and 4 weeks of treatment. fNIRS was used to detect the oxygenated hemoglobin values (HbO) in six ROIs at each time point. After treatment, both groups exhibited improvements in FM and MBI scores. The treatment group demonstrated significantly greater functional gains than the control group at both 2 and 4 weeks, as reflected in FM (T1T0: 5.867 ± 3.482 vs. 3.200 ± 2.077, P < 0.01, d = 0.93; T2T0: 13.533 ± 5.705 vs. 7.133 ± 2.503, P < 0.05, d = 1.45) and MBI scores (T1T0: 13.400 ± 7.129 vs. 8.133 ± 4.357, P < 0.05, d = 0.89; T2T0: 27.867 ± 10.106 vs. 16.467 ± 7.010, P < 0.05, d = 1.31). fNIRS data revealed that after 4 weeks, the treatment group showed significantly increased oxygenated hemoglobin levels in PMC and SMA compared to baseline (PMC: P < 0.001, d = 0.62; SMA: P < 0.001, d = 0.89), along with a more pronounced PMC activation and higher brain network efficiency relative to the control group (PMC: 0.019 ± 0.017 vs. 0.007 ± 0.005, P < 0.01, d = 1.01; network efficiency: P < 0.05). Moreover, improvements in brain network efficiency were positively correlated with gains in both FM and MBI scores across the cohort. Our study suggests that BCI treatment combined with conventional medical and rehabilitation therapy can effectively enhance motor function and activities of daily living in stroke patients with upper-limb dysfunction. Additionally, it can promote cortical activation in the ipsilesional PMC and SMA regions and improve the network efficiency between brain regions.