Transferring BCI models from calibration to control: Observing shifts in EEG features

TL;DR

A new paradigm containing a standard calibration session and a novel BCI control session based on EMG is demonstrated, which allows us to observe similarities in sensorimotor rhythms, and observe the additional preparation effects introduced by the control paradigm.

Abstract

Public Motor Imagery-based brain-computer interface (BCI) datasets are being used to develop increasingly good classifiers. However, they usually follow discrete paradigms where participants perform Motor Imagery at regularly timed intervals. It is often unclear what changes may happen in the EEG patterns when users attempt to perform a control task with such a BCI. This may lead to generalisation errors. We demonstrate a new paradigm containing a standard calibration session and a novel BCI control session based on EMG. This allows us to observe similarities in sensorimotor rhythms, and observe the additional preparation effects introduced by the control paradigm. In the Movement Related Cortical Potentials we found large differences between the calibration and control sessions. We demonstrate a CSP-based Machine Learning model trained on the calibration data that can make surprisingly good predictions on the BCI-controlled driving data.

Figures (8)

• Figure 1: Design of the data acquisition. First subjects do a calibration session following the Graz-BCI Motor Imagery paradigm. The EMG for this is used to develop a mock BCI, which the users then use to control a simulated car. The EMG, EEG and markers are recorded in both sessions resulting in a driving dataset and a calibration dataset.
• Figure 2: Timing of the calibration paradigm from OpenVibe.
• Figure 3: Participant view during the driving session.
• Figure 4: SMR during the left and right trials for calibration and driving. No baseline is used.
• Figure 5: Population averaged MRCP during calibration and driving. The movement starts at $t=1.25$, indicated by the rightmost dashed line. In the calibration plots the dashed line at $t=0$ shows when participants are given the directional cue. The blue line indicates the average ERP for left trials, the orange for right trials.