Abstract
BACKGROUND: Stroke is a significant cause of disability worldwide, often resulting in persistent upper-limb dysfunction. Robot-assisted therapy has emerged as an effective rehabilitation strategy by enabling intensive, repetitive, and task-specific training. In this study, we aimed to investigate the effects of resistance- versus assistance-based robotic interventions on brain activation and motor recovery in patients with stroke, as these effects remain insufficiently understood. METHODS: Twenty-five adults with hemiparetic stroke were randomized to a resistance-based robot training group (RTG, n = 13) or assistance-based robot training group (ATG, n = 12). Interventions were delivered using the InMotion 2.0 for 30 min per session, five sessions/week over 4 weeks. Brain activation was measured using functional near-infrared spectroscopy (fNIRS), motor function using the Fugl-Meyer Assessment for the Upper Limb (FM-UL) and kinematic indices from InMotion 2.0, and activities of daily living using the Motor Activity Log (MAL). RESULTS: Prefrontal activation decreased from pre- to post-intervention in both groups. In the ipsilesional hemispheres, differences between groups were significant (p < 0.05). In the resistance training group, additional improvements were found in mean velocity, circle size, and movement independence (p < 0.05). Both groups showed significant gains in FM-UL and ADL performance (p < 0.05), with no significant between-group differences in these measures. CONCLUSION: Resistance-based robotic training was associated with greater motor improvements in kinematic smoothness, and larger reductions in prefrontal activation within ipsilesional hemispheres compared with assistance-based training. These findings suggest differences in prefrontal activation patterns accompanied by improvements in kinematic movement smoothness in stroke survivors. CLINICAL TRIAL REGISTRATION: https://cris.nih.go.kr, (Registration Number: KCT0011076).