Assessing motor imagery in brain-computer interface training: Psychological and neurophysiological correlates. (March 2017)

• Record Type: Journal Article
• Authors: Vasilyev, Anatoly; Liburkina, Sofya; Yakovlev, Lev; Perepelkina, Olga; Kaplan, Alexander
• Is Part Of: Neuropsychologia. Volume 97(2017)
• Page Start: 56
• Page End: 65
• Publication Date: 2017-03
• Subjects: Motor imagery -- Brain-computer interface -- Transcranial magnetic stimulation -- Mu-rhythm -- Psychological assessment -- Neurorehabilitation; Neuropsychology -- Periodicals; Neurology -- Periodicals; Psychophysiology -- Periodicals; Neuropsychologie -- Périodiques; Neuropsychology; Periodicals; Electronic journals; 616.8
• DOI: 10.1016/j.neuropsychologia.2017.02.005
• Languages: English
• ISSNs: 0028-3932

Abstract

Abstract: Motor imagery (MI) is considered to be a promising cognitive tool for improving motor skills as well as for rehabilitation therapy of movement disorders. It is believed that MI training efficiency could be improved by using the brain-computer interface (BCI) technology providing real-time feedback on person's mental attempts. While BCI is indeed a convenient and motivating tool for practicing MI, it is not clear whether it could be used for predicting or measuring potential positive impact of the training. In this study, we are trying to establish whether the proficiency in BCI control is associated with any of the neurophysiological or psychological correlates of motor imagery, as well as to determine possible interrelations among them. For that purpose, we studied motor imagery in a group of 19 healthy BCI-trained volunteers and performed a correlation analysis across various quantitative assessment metrics. We examined subjects' sensorimotor event-related EEG events, corticospinal excitability changes estimated with single-pulse transcranial magnetic stimulation (TMS), BCI accuracy and self-assessment reports obtained with specially designed questionnaires and interview routine. Our results showed, expectedly, that BCI performance is dependent on the subject's capability to suppress EEG sensorimotor rhythms, which in turn is correlated with the idle state amplitude of those oscillations. Neither BCI accuracy nor the EEG features associated with MI were found to correlate with the level of corticospinal excitability increase during motor imagery, and with assessed imagery vividness. Finally, a significant correlation was found between the level of corticospinal excitability increase and kinesthetic vividness of imagery (KVIQ-20 questionnaire). Our results suggest that two distinct neurophysiological mechanisms might mediate possible effects of motor imagery: the non-specific cortical sensorimotor disinhibition and the focal corticospinal excitability increase. Acquired data suggests that BCI-based approach is unreliable in assessing motor imagery due to its high dependence on subject's innate EEG features (e.g. resting mu-rhythm amplitude). Therefore, employment of additional assessment protocols, such as TMS and psychological testing, is required for more comprehensive evaluation of the subject's motor imagery training efficiency. Highlights: EEG features do not predict cortical excitability changes during motor imagery. Kinesthetically vivid images produce the most effect on cortical excitability. EEG-based brain-computer interfaces are insufficient to assess motor imagery.