Effectiveness of Brain–Computer Interfaces with Biofeedback in the Rehabilitation of Cognitive Impairment after Stroke
摘要
Objective. To compare the effectivenesses of two brain–computer interface (BCI) software systems using biofeedback (BFB) and standard therapy in the restoration of cognitive functions after stroke. Materials and methods. The study involved 89 patients who had suffered stroke. Neuropsychological testing was performed using the Montreal Cognitive Assessment (MoCA) scale, the tracking test, subtest 9 of the Wechsler test, Kohs blocks, Schulte tables, and the A. R. Luriya 10 words memorization of test. Three groups were formed by simple randomization: the study group (n = 37), the reference group (n = 33), and the control group (n = 19). Subjects in the first group underwent BCI + BFB training based on the P300 rhythm; those in the second group underwent BCI + BFB training based on the EEG mu-rhythm, and those in the third group received standard therapy. Results. Increases in total MoCA scores were obtained in all three groups. Results in the first and second groups were comparable and exceeded those in the third group (p1–2 = 0.199, p1–3 < 0.001, p2–3 = 0.037). Effectiveness in the first group did not depend on the initial MoCA score, which was greater than that in the third group; in the second group, advantage over the third group was seen when initial MoCA scores were at least 22. Investigations using Schulte tables yielded comparable statistically significant changes in the first and second groups, as also occurred in the tracking test; there were no statistically significant changes in the control group. Investigations using Kohs blocks demonstrated greater statistically significant changes in the study group. The A. R. Luriya 10 words memorization test also showed more stable positive changes in the study group on analysis of memory disorders. Conclusions. The effectiveness of BCI + biofeedback was greater than that of standard therapy in poststroke cognitive impairments. The BCI + biofeedback used in the study group had the advantage over BCI + biofeedback in the reference group, which was associated with lower effectiveness of the latter when initial MoCA scores were <22 points, with lower effectiveness in the 10 words memorization and Kohs blocks tests.