Application of Brain-Computer Interactive Rehabilitation Training Combined With a Gait Robot: A Randomized Controlled Trial.

Abstract

To explore the effects of brain-computer interactive (BCI) rehabilitation training combined with gait robot (GR) on gait recovery in patients with hemiplegia after stroke. Randomized controlled trial. Hospital settings across the Shandong Provincial Third Hospital. A total of 120 eligible subjects were enrolled and randomly allocated, via random-number table, into 4 equal groups (n=30 each): control (routine training), BCI, GR, and BCI-GR (BCI combined with GR training). Each group received its designated training once daily, 6 d/wk, for 8 consecutive weeks. Fugl-Meyer Assessment-Lower Extremity (FMA-LE), Berg Balance Scale (BBS), Functional Ambulation Category (FAC), integrated electromyography (IEMG), and stride parameters. Assessments were conducted at baseline, 4 wk, and 8 wk. After 8 wk, all interventions-BCI, GR, and BCI-GR-significantly improved lower-limb function, muscle activity, and gait compared with control (all P<.01). FMA-LE increased by +12.93 (BCI-GR), +12.16 (BCI), and +12.07 (GR) versus +8.20 in control; BBS improved by +14.53, +13.63, and +14.30 versus +9.53; FAC improved by +1.73, +1.73, and +1.77 versus +1.37. IEMG of tibialis anterior and gastrocnemius increased most in BCI-GR (+0.156 and +0.063), significantly higher than BCI and GR (P<.05), whereas co-contraction ratio decreased most in BCI-GR (-18.59%; P<.05). Stride parameters (step frequency, step length, step width, and walking speed) improved in all intervention groups, with BCI-GR showing greater gains in step frequency (+25.90) and step width reduction (-12.87) versus single interventions (P<.05). No other significant differences among intervention groups were observed at week 8. BCI and GR interventions significantly improve lower-limb motor function, muscle activation, and gait in poststroke patients. Combined BCI-GR training accelerates early functional recovery and more effectively enhances muscle activation patterns compared with single interventions, providing a promising strategy for poststroke rehabilitation.