F191. Electrophysiological brain activity during motor imagery enhanced by brain–computer interface in healthy volunteers and post-stroke patients. (May 2018)

• Record Type: Journal Article
• Authors: Aziatskaya, Guzel; Lyukmanov, Roman; Frolov, Aleksandr; Bobrov, Pavel; Fedotova, Irina; Husek, Dusan; Snasel, Vaclav; Suponeva, Natalia; Piradov, Michael; Poydasheva, Alexandra
• Is Part Of: Clinical neurophysiology. Volume 129(2018)Supplement 1
• Page Start: e140
• Publication Date: 2018-05
• Subjects: Neurophysiology -- Periodicals; Electroencephalography -- Periodicals; Electromyography -- Periodicals; Neurology -- Periodicals; 612.8
• DOI: 10.1016/j.clinph.2018.04.354
• Languages: English
• ISSNs: 1388-2457

Abstract

Abstract : Introduction: Motor imagery (MI) has been shown to be an effective tool in rehabilitation of post-stroke patients. MI can be enhanced by kinesthetic feedback provided by brain-computer-exoskeleton interfaces. Description and analysis of brain activity during the kinesthetic imagery of movements in healthy subjects and post-stroke patients with the subcortical location of the damage was performed in the current study. Methods: Seven right-handed healthy volunteers (age 29 [22.0; 35.0] years) and nine post-stroke patients (age 52.0 [36.0; 58.0] years, time since stroke 6.0 [4.0; 13.0] months) were enrolled. They were trained in kinesthetic motor imagery using brain-computer interface (BCI) and a controllable exoskeleton. All participants had 20 training sessions with a duration of up to 15 min. The BCI was controlled using a Bayessian classifier, which was preset to recognize the imagery of left or right hand extension and the rest state. Processing of EEG recordings was performed using the independent component analysis. The digitization of EEG-electrode positions for obtaining their coordinates in the MRI-related system was made using an NBS Eximia Nexstim device for navigational transcranial magnetic stimulation. This study used the AMICA method to find the sources of electrophysiological brain activity during motor imagery. For each component determined, the EEG inverse problem was solved according to the sLORETA method. Results: Components that are most frequently identified by the AMICA method were revealed in the same areas in both groups: SIL (the primary somatosensory left cortex), SIR (the primary somatosensory right cortex), PRC (the precuneus), PRM (the premotor left cortex), and SMA (the supplementary motor area). There were differences in the activity characteristics when comparing healthy controls and patients: the μ -rhythm in the somatosensory area was characterized by a lower frequency in the studied poststroke patients than in the healthy subjects; it had a poorly pronounced peak in all mental states and was only slightly suppressed during contralateral hand motor imagery. The rhythm in the α -band was also characterized by a low frequency with a poorly pronounced peak in the precuneus and premotor area; however, unlike in the healthy subjects, this rhythm increased, i.e., was synchronized, during motor imagery. Conclusion: This study is an extension of previous investigations of EEGs during motor imagery in healthy subjects. Only two symmetrical sources in the primary somatosensory cortex were previously discussed. This study revealed three more sources: the premotor area, precuneus, and supplementary motor area. The characteristics of the EEG sources during motor imagery are fundamentally different in post-stroke patients comparing to healthy controls. Funding was provided by grant RFMEF N I60715X0128.