Riemannian Channel Selection for BCI With Between-Session Non-Stationarity Reduction Capabilities

Abstract

Between-session non-stationarity is a major challenge of current Brain-Computer Interfaces (BCIs) that affects system performance. In this paper, we investigate the use of channel selection for reducing between-session non-stationarity with Riemannian BCI classifiers. We use the Riemannian geometry framework of covariance matrices due to its robustness and promising performances. Current Riemannian channel selection methods do not consider between-session non-stationarity and are usually tested on a single session. Here, we propose a new channel selection approach that specifically considers non-stationarity effects and is assessed on multi-session BCI data sets. Methods: We remove the least significant channels using a sequential floating backward selection search strategy. Our contributions include: 1) quantifying the non-stationarity effects on brain activity in multi-class problems by different criteria in a Riemannian framework and 2) a method to predict whether BCI performance can improve using channel selection. Results: We evaluate the proposed approaches on three multisession and multi-class mental tasks (MT)-based BCI datasets. They could lead to significant improvements in performance as compared to using all channels for datasets affected by between-session non-stationarity and to significant superiority to the state-of-the-art Riemannian channel selection methods over all datasets, notably when selecting small channel set sizes. Conclusion: Reducing non-stationarity by channel selection could significantly improve Riemannian BCI classification accuracy. Significance: Our proposed channel selection approach contributes to make Riemannian BCI classifiers more robust to between-session non-stationarities.

Enhancing Cross-Subject Motor Imagery Classification in EEG-Based Brain–Computer Interfaces by Using Multi-Branch CNN

“…There may be further opportunities to explore more multi-branch designs. As EEG data is non-stationary, which has implications for the feature extraction process, different approaches can be explored to address the issue of non-stationarity [ 51 , 52 , 53 ]. Transfer learning approaches can be implemented to resolve the issue of less data per subject and inter-subject variations of the EEG-based dataset.…”

An EEG-Based Brain-Computer Interface Using Spectral Correlation Function

“…As pointed out by Sadatnejad et al [73], in the current BCI systems, between-session non-stationarity poses a significant performance issue. They investigated the application of the channel selection technique with Riemannian BCI classifiers to reduce between-session nonstationarity.…”

Novel SPD Matrix Representations Considering Cross-Frequency Coupling for EEG Classification Using Riemannian Geometry

“…To further improve BCIs, various Riemannnian specific approaches have been developed including EEG channel selection [8], dimensionality reduction [9], artifact/outlier detection [10] or frequency band (FB) selection [11], among other. All these methods have in common that they operate after the covariance matrix estimations, and all apply on conventional cross-channel Cov only.…”