A learnable EEG channel selection method for MI-BCI using efficient channel attention.

Electroencephalography (EEG) stands as a noninvasive and cost-effective method for recording neural activity, holding potential for applications such as identifying neural processes underlying human emotions. This paper delves into the transferability and generalizability of EEG channel selection in emotion recognition, adopting a dataset-independent approach. By leveraging Power Spectral Density (PSD), we identify high-contributing EEG channels in the SEED V dataset and validate our approach on the independent SEED IV dataset using a Convolutional Neural Network (CNN) model. The channel selection method helped in eliminating insignificant EEG channels, which can improve the applicability of developing more efficient EEG devices for daily use to monitor emotions, as well as in individuals suffering from various neurodegenerative diseases. Through extensive experiments varying the number of channels and features, our model achieves classification accuracies of 77.02%, 75.42%, 71.31%, and 64.31% with 62, 30, 20, and 10 EEG channels, accompanied by 310, 90, 60, and 30 Differential Entropy (DE) features respectively. Further, the proposed approach is tested by introducing Gaussian noise to the training set and evaluating its sensitivity to signal noise. Finally, results are compared with state-of-the-art models highlighting the potential of our dataset-independent channel selection method.

Objective. Channel selection in the electroencephalogram (EEG)-based brain–computer interface (BCI) has been extensively studied for over two decades, with the goal being to select optimal subject-specific channels that can enhance the overall decoding efficacy of the BCI. With the emergence of deep learning (DL)-based BCI models, there arises a need for fresh perspectives and novel techniques to conduct channel selection. In this regard, subject-independent channel selection is relevant, since DL models trained using cross-subject data offer superior performance, and the impact of inherent inter-subject variability of EEG characteristics on subject-independent DL training is not yet fully understood. Approach. Here, we propose a novel methodology for implementing subject-independent channel selection in DL-based motor imagery (MI)-BCI, using layer-wise relevance propagation (LRP) and neural network pruning. Experiments were conducted using Deep ConvNet and 62-channel MI data from the Korea University EEG dataset. Main Results. Using our proposed methodology, we achieved a 61% reduction in the number of channels without any significant drop (p = 0.09) in subject-independent classification accuracy, due to the selection of highly relevant channels by LRP. LRP relevance-based channel selections provide significantly better accuracies compared to conventional weight-based selections while using less than 40% of the total number of channels, with differences in accuracies ranging from 5.96% to 1.72%. The performance of the adapted sparse-LRP model using only 16% of the total number of channels is similar to that of the adapted baseline model (p = 0.13). Furthermore, the accuracy of the adapted sparse-LRP model using only 35% of the total number of channels exceeded that of the adapted baseline model by 0.53% (p = 0.81). Analyses of channels chosen by LRP confirm the neurophysiological plausibility of selection, and emphasize the influence of motor, parietal, and occipital channels in MI-EEG classification. Significance. The proposed method addresses a traditional issue in EEG-BCI decoding, while being relevant and applicable to the latest developments in the field of BCI. We believe that our work brings forth an interesting and important application of model interpretability as a problem-solving technique.

With the development of deep learning, many motor imagery brain-computer interfaces based on convolutional neural networks (CNNs) show outstanding performances. However, the trial number of EEG in the training set is usually limited, and redundancy extensively exists in multiple channel EEG. Thus, overfitting often appears in CNN based motor imagery recognition model and greatly affects the performances of model. In this paper, channel drop out is proposed to address this problem by data augmentation and ensemble learning. Specifically, one of all EEG channels will be dropped and replaced by the mean signal of all EEG channels. In this way, the trial number in the training set was enlarged by channel drop out. And at the testing stage, all the EEG trials processed by channel drop out were fed to the CNN model and the average output probabilities of them were applied to determine the prediction. The experiments were conducted on two popular CNN models applied in motor imagery BCI and BCI Competition IV datasets 2a to verify the performances of the proposed channel drop out approach. The results show that average improvements provided by channel drop out in two-category or four-category motor imagery classification are 2.83% and 2.65% compared with the original CNN model. So the channel drop out approach significantly improves the performances of motor imagery based BCI.

In recent years, attention mechanisms have shown great potential in various computer vision tasks. However, most existing methods focus on developing more complex attention modules for better performance, which inevitably increases the complexity of the model. To overcome performance and complexity tradeoffs, this paper proposes efficient spatial and channel attention (ESCA), a symmetric, comprehensive, and efficient attention module. By analyzing squeeze-and-excitation (SE), convolutional block attention module (CBAM), coordinate attention (CA), and efficient channel attention (ECA) modules, we abandon the dimension-reduction operation of SE module, verify the negative impact of global max pooling (GMP) on the model, and apply a local cross-channel interaction strategy without dimension reduction to learn attention. We not only care about the channel features of the image, we also care about the spatial location of the target on the image, and we take into account the effectiveness of channel attention, so we designed the symmetric ESCA module. The ESCA module is effective, as demonstrated by its application in the ResNet-50 classification benchmark. With 26.26 M parameters and 8.545 G FLOPs, it introduces a mere 0.14% increment in FLOPs while achieving over 6.33% improvement in Top-1 accuracy and exceeding 3.25% gain in Top-5 accuracy. We perform image classification and object detection tasks on ResNet, MobileNet, YOLO, and other architectures on popular datasets such as Mini ImageNet, CIFAR-10, and VOC 2007. Experiments show that ESCA can achieve great improvement in model accuracy at a very small cost, and it performs well among similar models.

This article proposes an access-point and channel selection method for Internet of Things environments. Recently, the number of wireless nodes has increased with the growth of Internet of Things technologies. In order to accommodate traffic generated by the wireless nodes, we need to utilize densely placed wireless access-points. This article introduces a joint optimization problem of access-point and channel selection for such an environment. The proposed method deals with the optimization problem, using Markov approximation which adapts to dynamic changes in network conditions. Markov approximation is a distributed optimization framework, where a network is optimized by individual behavior of users forming a time-reversible continuous-time Markov chain. The proposed method searches optimal solution for the access-point and channel selection problem on the time-reversible continuous-time Markov chain. Simulation experiments demonstrate the effectiveness of the proposed method.

The Deep Learning (DL) approach has been gaining much popularity in recent years in the development of electroencephalogram (EEG) based Motor Imagery (MI) Brain-Computer Interface (BCI) systems, aiming to improve the performance of existing stroke rehabilitation strategies. A complex deep neural network structure has lots of neurons with thousands of parameters to optimize, and a great deal of data is often required to train the network and the training process can take an extremely long time. High training costs and high model complexity not only have negative impacts on the performance of the BCI system but also on its applicability to meet the real-time requirement to support the rehabilitation exercises of patients. To tackle the challenge, a contribution-based neuron selection method is proposed in this paper. A Convolutional Neural Network (CNN) based motor imagery classification framework is implemented, and a neuron pruning approach is developed and applied. The temporal and spatial features of EEG signals are captured by the CNN layers, and then the fast recursive algorithm (FRA) is applied to prune redundant parameters in the fully connected layers which reduces the computation cost of the CNN model without affecting its performance. The experimental results show that the proposed method can achieve up to 50% model size reduction and 67.09% computation savings.

Although it is clinically important, a reliable and economical solution to automatic seizure detection for patients at home is yet to be developed. Traditional algorithms rely on multi-channel EEG signals and features of canonical EEG power description. This study is aimed to propose an effective single-channel EEG seizure detection method centered on novel EEG power parameterization and channel selection algorithms.We employed the publicly available multi-channel CHB-MIT Scalp EEG database to gauge the effectiveness of our approach. We first adapted a power spectra parameterization algorithm to characterize the aperiodic and periodic components of the ictal and inter-ictal EEGs. We selected four features based on their statistical significance and interpretability, and developed a ranking approach to channel selection for each patient. We then tested the effectiveness of our approaches to channel and feature selection for automatic seizure detection using support vector machine (SVM) as the classifier. The performance of our algorithm was evaluated using five-fold cross-validation and compared to those methods of comparable complexity (using one or two channels of EEG), in terms of accuracy, specificity, sensitivity, precision and F1 score.Some channels of EEG signals show strikingly different distributions of PSD features between the ictal and inter-ictal states. Four features including the offset and exponent parameters for the aperiodic component and the first and second highest total power (TPW1 and TPW2) form the basis of channel selection and the input of SVM classifier. The selected channel is found to be patient-specific. Our approach has achieved a mean sensitivity of 95.6%, specificity of 99.2%, accuracy of 98.6%, precision of 95.5%, and F1 score of 95.5%. Compared with algorithms in previous studies that used one or two channels of EEG signals, ours outperforms in specificity and accuracy with comparable sensitivity.EEG power spectra parameterization to feature extraction and feature ranking-based channel selection are found to enable efficient and effective automatic seizure detection based on single-channel EEG signal.

How many channels are suitable for independent component analysis in motor imagery brain-computer interface

Convolution neural networks (CNNs) have been achieving increasing attention for the artificial bandwidth extension (ABE) task recently. However, these methods use the flipped low-frequency phase to reconstruct speech signals, which may lead to the well-known invalid short-time Fourier Transform (STFT) problem. The convolutional operations only enable networks to construct informative features by fusing both channel-wise and spatial information within local receptive fields at each layer. In this paper, we introduce a Time-Frequency Network (TFNet) with channel attention (CA) and non-local (NL) modules for ABE. The TFNet exploits the information from both time and frequency domain branches concurrently to avoid the invalid STFT problem. To capture the channels and space dependencies, we incorporate the CA and NL modules to construct a proposed fully convolutional neural network for the time and frequency branches of TFNet. Experimental results demonstrate that the proposed method outperforms the competing method.

In the field of brain-computer interface (BCI) based on motor imagery (MI), multi-channel electroencephalography (EEG) data is commonly utilized for MI task recognition to achieve sensory compensation or precise human-computer interaction. However, individual physiological differences, environmental variations, or redundant information and noise in certain channels can pose challenges and impact the performance of BCI systems. In this study, we introduce a channel selection method utilizing Hybrid-Recursive Feature Elimination (H-RFE) combined with residual graph neural networks for MI recognition. This channel selection method employs a recursive feature elimination strategy and integrates three classification methods, namely random forest, gradient boosting, and logistic regression, as evaluators for adaptive channel selection tailored to specific subjects. To fully exploit the spatiotemporal information of multi-channel EEG, this study employed a graph neural network embedded with residual blocks to achieve precise recognition of motor imagery. We conducted algorithm testing using the SHU dataset and the PhysioNet dataset. Experimental results show that on the SHU dataset, utilizing 73.44% of the total channels, the cross-session MI recognition accuracy is 90.03%. Similarly, in the PhysioNet dataset, using 72.5% of the channel data, the classification result also reaches 93.99%. Compared to traditional strategies such as selecting three specific channels, correlation-based channel selection, mutual information-based channel selection, and adaptive channel selection based on Pearson coefficients and spatial positions, the proposed method improved classification accuracy by 34.64%, 10.8%, 3.25% and 2.88% on the SHU dataset, and by 46.96%, 5.04%, 5.81% and 2.32% on the PhysioNet dataset, respectively.

Background Due to the redundant information contained in multichannel electroencephalogram (EEG) signals, the classification accuracy of brain-computer interface (BCI) systems may deteriorate to a large extent. Channel selection methods can help to remove task-independent electroencephalogram (EEG) signals and hence improve the performance of BCI systems. However, in different frequency bands, brain areas associated with motor imagery are not exactly the same, which will result in the inability of traditional channel selection methods to extract effective EEG features. New Method To address the above problem, this paper proposes a novel method based on common spatial pattern- (CSP-) rank channel selection for multifrequency band EEG (CSP-R-MF). It combines the multiband signal decomposition filtering and the CSP-rank channel selection methods to select significant channels, and then linear discriminant analysis (LDA) was used to calculate the classification accuracy. Results The results showed that our proposed CSP-R-MF method could significantly improve the average classification accuracy compared with the CSP-rank channel selection method.

This paper presents an automatic selected channel method for improving brain-computer interface (BCI) performance. Finding an effective channel composition in a BCI system is important for parsing the complexity of these systems. In this study, several statistical methods based on an energy extraction method were used to automatically detect the composition of selected channels. We introduce four techniques for automatic channel selection to optimize the energy extraction method, such as mean, high mean, box, and high in box. The performance of all of the proposed techniques was evaluated based on performance accuracy, compression ratio, and channel mapping on a brain scalp. The performance of the proposed BCI framework was compared against a BCI framework that used a manual technique as the energy extraction method. The proposed BCI framework system used a conventional common spatial pattern (CSP) to extract features from two-class motor imagery EEG signals before employing extreme learning machine (ELM) to classify the features of the EEG signal. As a result, the proposed automatic channel selection methods were found effective in finding optimal channels and provided better performance accuracy. In general, the proposed method improved the conventional BCI performance by up to 16% accuracy and 87% channel compression size. Besides that, the automatic technique also yielded better BCI performance accuracy of up to 5% compared to the BCI system that used the manual technique of energy extraction as its EEG channel selection method.

BackgroundThe aging of the world population poses a major health challenge, and brain–computer interface (BCI) technology has the potential to provide assistance and rehabilitation for the elderly.ObjectivesThis study aimed to investigate the electroencephalogram (EEG) characteristics during motor imagery by comparing young and elderly, and study Convolutional Neural Networks (CNNs) classification for the elderly population in terms of fatigue analysis in both frontal and parietal regions.MethodsA total of 20 healthy individuals participated in the study, including 10 young and 10 older adults. All participants completed the left- and right-hand motor imagery experiment. The energy changes in the motor imagery process were analyzed using time–frequency graphs and quantified event-related desynchronization (ERD) values. The fatigue level of the motor imagery was assessed by two indicators: (θ + α)/β and θ/β, and fatigue-sensitive channels were distinguished from the parietal region of the brain. Then, rhythm entropy was introduced to analyze the complexity of the cognitive activity. The phase-lock values related to the parietal and frontal lobes were calculated, and their temporal synchronization was discussed. Finally, the motor imagery EEG data was classified by CNNs, and the accuracy was discussed based on the analysis results.ResultFor the young and elderly, ERD was observed in C3 and C4 channels, and their fatigue-sensitive channels in the parietal region were slightly different. During the experiment, the rhythm entropy of the frontal lobe showed a decreasing trend with time for most of the young subjects, while there was an increasing trend for most of the older ones. Using the CNN classification method, the elderly achieved around 70% of the average classification accuracy, which is almost the same for the young adults.ConclusionCompared with the young adults, the elderly are less affected by the level of cognitive fatigue during motor imagery, but the classification accuracy of motor imagery data in the elderly may be slightly lower than that in young persons. At the same time, the deep learning method also provides a potentially feasible option for the application of motor-imagery BCI (MI-BCI) in the elderly by considering the ERD and fatigue phenomenon together.

Lane detection based on semantic segmentation can achieve high accuracy, but, in recent years, it does not have a mobile-friendly cost, which is caused by the complex iteration and costly convolutions in convolutional neural networks (CNNs) and state-of-the-art (SOTA) models based on CNNs, such as spatial CNNs (SCNNs). Although the SCNN has shown its capacity to capture the spatial relationships of pixels across rows and columns of an image, the computational cost and memory requirement needed cannot be afforded with mobile lane detection. Inspired by the channel attention and self-attention machine, we propose an integrated coordinate attention (ICA) module to capture the spatial-wise relationships of pixels. Furthermore, due to the lack of enhancement in the channel dimension, we created an efficient network with a channel-enhanced coordinate attention block named CCA, composed of ICA and other channel attention modules, for all-dimension feature enhancement. As a result, by replacing many repeated or iterative convolutions with the attention mechanism, CCA reduces the computational complexity. Thus, our method achieves a balance of accuracy and speed and has better performance on two lane datasets—TuSimple and ILane. At less than a few tenths of the computational cost, our CCA achieves superior accuracy compared to the SCNN. These results show that the low cost and great performance of our design enable the use of the lane detection task in autopilot scenarios.

Brain computer interface (BCI) is a technology recording brain signals and translate into machine language to drive device movement. To improve the convenience and practicability of BCI systems, it is necessary to reduce the number of channels of EEG without reducing classification accuracy. In this paper, a channel selection method was proposed. Combine continuous wavelet transform (CWT) with EEG source imaging (ESI), we can find out the source distribution of two kinds of motor imagery (MI) tasks on the cortex. Then the different value of source distribution on the cortex was projected to scalp. The value on the scalp was treated as a criterion for channel selection. While ESI was applied for locating the active areas, we can not only find out the prior channels clearly, but also figure out the expected areas briefly. The satisfying results show that the combination of CWT and ESI is a potential method for dealing with MI tasks.