Neurophysiological predictor of SMR-based BCI performance

Abstract

Brain-Computer Interfaces (BCIs) allow a user to control a computer application by brain activity as measured, e.g., by electroencephalography (EEG).After about 30 years of BCI research, the success of control that is achieved by means of a BCI system still greatly varies between subjects. For about 20% of potential users the obtained accuracy does not reach the level criterion, meaning that BCI control is not accurate enough to control an application.The determination of factors that may serve to predict BCI performance, and the development of methods to quantify a predictor value from psychological and/or physiological data serves two purposes: a better understanding of the 'BCI-illiteracy phenomenon', and avoidance of a costly and eventually * Corresponding author. Berlin Institute of Technology, Machine Learning Laboratory, Sekr. FR6-9, Franklinstrasse 28/29, 10587 Berlin, Germany. Tel: +49 3031478624, Fax: +49 3031478622. E-mail address: dickhaus@cs.tu-berlin.de (Th. Dickhaus). Preprint submitted to NeuroImageMarch 8 which operates on modulations of sensory motor rhythms (SMRs).

Using a motor imagery questionnaire to estimate the performance of a Brain–Computer Interface based on object oriented motor imagery

“…Another result from this study, namely that people with a lower KI score have a lower intensity of the SMI peak, is in accordance with a study by Blankertz et al [2010]. In that study correlation between the power amplitude of the SMR over the C3+C4 area of the cortex and the classification accuracy of a BCI classifier based on ERS/ERD was demonstrated, showing that people with a lower SMR in a relaxed state exhibit worse BCI performance.…”

“…There have been several studies attempting to define suitable candidates for MI based BCI. Blankertz et al [2010] suggested recording a brief EEG session in a relaxed state with eyes open, to detect power amplitude of SMR in the 'idle' state. They showed good correlation between the power amplitude and subsequent BCI classification accuracy in subjects performing cue-based imagination with feedback.…”

Closed-loop brain training: the science of neurofeedback

“…Clinical phenotyping may help to identify responders, and neurofeedback protocols could be designed based on network models of neural dysfunction 147 (BOX 2) rather than on patient interviews and self-reports as in current practice 148 . Finally, training protocols could be tailored to individual patients based on predictions of neurofeedback performance from resting-state activity 149 , anatomical brain structure 150,151 and personality traits 152 .…”

Common Spatial Pattern Patches: Online evaluation on BCI-naive users

“…Run 1: subject-independent 1) Pre-trained CSPP and LDA: For each binary combination of motor imagery classes, CSPP filters were calculated from a data base of 48 successful BCI performers (data described in [11]). The data were concatenated, band-pass filtered in a broad band 8-32 Hz (previously chosen by offline analyses) and epoched in the time interval 750-3750 ms after stimulus onset.…”

“…Brain Computer Interfaces (BCI) based on sensorimotor rhythms (SMRs), make use of the voluntary modulation in the alpha (8)(9)(10)(11)(12) or beta (13-20 Hz) frequency band of the electroencephalography (EEG) activity over sensorimotor areas during limb movement imagination to obtain, by proper real time processing of the brain activity, a control signal for an external device. Despite a great progress in BCI research (see [1] for a review), still 20-30% of the healthy population is not able to reach the level criterion of 70% of accuracy, above which users feel to have obtained BCI control, as assessed in [2] by a psychological study for two-class BCI with communication applications.…”