An automatic channel selection method based on the standard deviation of wavelet coefficients for motor imagery based brain–computer interfacing

Abstract

The redundant data in multichannel electroencephalogram (EEG) signals significantly reduces the performance of brain–computer interface (BCI) systems. By removing redundant channels, a channel selection strategy increases the classification accuracy of BCI systems. In this work, a novel channel selection method (stdWC) based on the standard deviation of wavelet coefficients across channels is proposed to identify Motor Imagery (MI) based EEG signals. The wavelet coefficients are calculated by employing a Continuous Wavelet Transform (CWT) filter bank to decompose each trial from the EEG channel. The wavelet coefficient's standard deviation values are obtained across the channels, and these values are then sorted to determine the EEG channels with the highest standard deviation values. The channels with the largest wavelet coefficient divergence are chosen. MI trials are then spatially filtered with the Common Spatial Pattern (CSP), and CWT filter bank‐based 2D images are generated from the spatially filtered trials. These images are then classified using a unique nine‐layered convolutional neural network (CNN) model that combines two feature maps acquired with differing filter sizes. The proposed framework (stdWC‐CSP‐CNN) is evaluated using kappa score and classification accuracy on two publically accessible datasets (BCI Competition III dataset IVa and BCI Competition IV dataset 2a). The suggested framework achieved a mean test classification accuracy of 88.8% for dataset IVa from BCI Competition III and 75.03% for dataset 2a from BCI Competition IV, according to the results. The proposed channel selection method outperforms the other channel selection methods examined, according to the results. By rejecting redundant channels, the whole framework can improve the performance of MI‐based BCIs.

Optimal Channel Selection of Multiclass Motor Imagery Classification Based on Fusion Convolutional Neural Network with Attention Blocks

“…Different methods for selecting channels have been used with the BCI IV 2a dataset, specifically when dealing with four-class classification. Researchers in [18][19][20][21][22] employ three main classification techniques: one-vs-one, one-vs-rest, and multiclass classification. In both one-vs-one and one-vs-rest, the means are derived from multiple binary classifications.…”

“…They demonstrated the advantages of incorporating feature extraction, feature selection, and MDA-SOGWO channel selection to enhance classification accuracy, elevating it from 67.04% to 80.82%. Additionally, the authors of [18,21] demonstrate the use of DL classification with the one-vs-rest strategy to validate channel selection methods. In their work [18], the researchers employed CSPs for optimal channel selection, followed by Fast Fourier Transform (FFT) transformation before training the DL model.…”

“…As a result of channel selection, the accuracy of the SSAE model increased to 71.31%, while that of the DBN-RBM model increased to 68.63%. As detailed in [21], the authors developed a channel selection approach based on the standard deviation of wavelet coefficients across channels. They implemented CSPs using a one-vs-rest strategy and then utilized a CNN model for data classification, achieving an accuracy of 75.03%.…”

“…Based on the channel selection process, when selecting the optimal channels, it is possible to improve the classification results while ensuring that no significant channels are removed [3]. Due to this, the majority of researchers [18][19][20][21][22] selected different channels for each subject individually. Selecting a fixed channel for all subjects is a problem that has not previously been addressed.…”

“…As an example, let us consider that we have these two fitness parents Parent1 = [1, 3,7,15,19,22] and Parent2 = [2, 7,15,21]. Assume that p1 = 3, p2 = 2, p3 = 2, and p4 = 4.…”