Intra- and Inter-subject Variability in EEG-Based Sensorimotor Brain Computer Interface: A Review

• Published in: Frontiers in computational neuroscience Vol. 13; p. 87
• Main Authors: Saha, Simanto, Baumert, Mathias
• Format: Journal Article
• Language: English
• Published: Switzerland Frontiers Research Foundation 21.01.2020; Frontiers Media S.A
• Subjects: Brain; brain computer interface; Computer applications; EEG; Electroencephalography; Implants; inter-subject associativity; Interfaces; Learning; Learning algorithms; Machine learning; Neurophysiology; Neuroscience; Prostheses; Rehabilitation; sensorimotor rhythms; Sensorimotor system; Stroke; Topography; Transfer learning; brain computer interface; electroencephalography; transfer learning; inter-subject associativity; sensorimotor rhythms
• ISSN: 1662-5188; 1662-5188
• DOI: 10.3389/fncom.2019.00087

Brain computer interfaces (BCI) for the rehabilitation of motor impairments exploit sensorimotor rhythms (SMR) in the electroencephalogram (EEG). However, the neurophysiological processes underpinning the SMR often vary over time and across subjects. Inherent intra- and inter-subject variability causes covariate shift in data distributions that impede the transferability of model parameters amongst sessions/subjects. Transfer learning includes machine learning-based methods to compensate for inter-subject and inter-session (intra-subject) variability manifested in EEG-derived feature distributions as a covariate shift for BCI. Besides transfer learning approaches, recent studies have explored psychological and neurophysiological predictors as well as inter-subject associativity assessment, which may augment transfer learning in EEG-based BCI. Here, we highlight the importance of measuring inter-session/subject performance predictors for generalized BCI frameworks for both normal and motor-impaired people, reducing the necessity for tedious and annoying calibration sessions and BCI training.