A new automated multi-stage system of non-local means and multi-kernel adaptive filtering techniques for EEG noise and artifacts suppression

FLANet: A multiscale temporal convolution and spatial-spectral attention network for EEG artifact removal with adversarial training

Objective.Denoising artifacts, such as noise from muscle or cardiac activity, is a crucial and ubiquitous concern in neurophysiological signal processing, particularly for enhancing the signal-to-noise ratio in electroencephalograph (EEG) analysis. Novel methods based on deep learning demonstrate a notably prominent effect compared to traditional denoising approaches. However, those still suffer from certain limitations. Some methods often neglect the multi-domain characteristics of the artifact signal. Even among those that do consider these, there are deficiencies in terms of efficiency, effectiveness and computation cost.Approach.In this study, we propose a multiscale temporal convolution and spatial-spectral attention network with adversarial training for automatically filtering artifacts, named filter artifacts network (FLANet). The multiscale convolution module can extract sufficient temporal information and the spatial-spectral attention network can extract not only non-local similarity but also spectral dependencies. To make data denoising more efficient and accurate, we adopt adversarial training with novel loss functions to generate outputs that are closer to pure signals.Main results.The results show that the method proposed in this paper achieves great performance in artifact removal and valid information preservation on EEG signals contaminated by different types of artifacts. This approach enables a more optimal trade-off between denoising efficacy and computational overhead.Significance.The proposed artifact removal framework facilitates the implementation of an efficient denoising method, contributing to the advancement of neural analysis and neural engineering, and can be expected to be applied to clinical research and to realize novel human-computer interaction systems.

A regression model combined convolutional neural network and recurrent neural network for electroencephalogram-based cross-subject fatigue detection

Fatigue, one of the most important factors affecting road safety, has attracted many researchers' attention. Most existing fatigue detection methods are based on feature engineering and classification models. The feature engineering is greatly influenced by researchers' domain knowledge, which will lead to a poor performance in fatigue detection, especially in cross-subject experiment design. In addition, fatigue detection is often simplified as a classification problem of several discrete states. Models based on deep learning can realize automatic feature extraction without the limitation of researcher's domain knowledge. Therefore, this paper proposes a regression model combined convolutional neural network and recurrent neural network for electroencephalogram-based (EEG-based) cross-subject fatigue detection. At the same time, a twofold random-offset zero-overlapping sampling method is proposed to train a bigger model and reduce overfitting. Compared with existing results, the proposed method achieves a much better result of 0.94 correlation coefficient (COR) and 0.09 root mean square error (RMSE) in a within-subject experiment design. What is more, there is no misclassification between awake and drowsy states. For cross-subject experiment design, the COR and RMSE are 0.79 and 0.15, respectively, which are close to the existing within-subject results and better than similar cross-subject results. The cross-subject regression model is very important for fatigue detection application since the fatigue indication is more precise than several discrete states and no model calibration is required for a new user. The twofold random-offset zero-overlapping sampling method can also be used as a reference by other EEG-based deep learning research.

Identification and classification of epileptic EEG signals using invertible constant-Qtransform-based deep convolutional neural network

Context.Epilepsy is the most widespread disorder of the nervous system, affecting humans of all ages and races. The most common diagnostic test in epilepsy is the electroencephalography (EEG).Objective.In this paper, a novel automated deep learning approach based on integrating a pre-trained convolutional neural network (CNN) structure, called AlexNet, with the constant-Qnon-stationary Gabor transform (CQ-NSGT) algorithm is proposed for classifying seizure versus seizure-free EEG records.Approach.The CQ-NSGT method is introduced to transform the input 1D EEG signal into 2D spectrogram which is sent to the AlexNet CNN model. The AlexNet architecture is utilized to capture the discriminating features of the 2D image corresponding to each EEG signal in order to distinguish seizure and non-seizure subjects using multi-layer perceptron algorithm.Main results. The robustness of the introduced CQ-NSGT technique in transforming the 1D EEG signals into 2D spectrograms is assessed by comparing its classification results with the continuous wavelet transform method, and the results elucidate the high performance of the CQ-NSGT technique. The suggested epileptic seizure classification framework is investigated with clinical EEG data acquired from the Bonn University database, and the experimental results reveal the superior performance of the proposed framework over other state-of-the-art approaches with an accuracy of 99.56%, sensitivity of 99.12%, specificity of 99.67%, and precision of 98.69%.Significance.This elucidates the importance of the proposed automated system in helping neurologists to accurately interpret and classify epileptic EEG records without necessitating tedious visual inspection or massive data analysis for long-term EEG signals.

EEGDnet: Fusing non-local and local self-similarity for EEG signal denoising with transformer

Electroencephalogram (EEG) has shown a useful approach to produce a brain-computer interface (BCI). One-dimensional (1-D) EEG signal is yet easily disturbed by certain artifacts (a.k.a. noise) due to the high temporal resolution. Thus, it is crucial to remove the noise in received EEG signal. Recently, deep learning-based EEG signal denoising approaches have achieved impressive performance compared with traditional ones. It is well known that the characteristics of self-similarity (including non-local and local ones) of data (e.g., natural images and time-domain signals) are widely leveraged for denoising. However, existing deep learning-based EEG signal denoising methods ignore either the non-local self-similarity (e.g., 1-D convolutional neural network) or local one (e.g., fully connected network and recurrent neural network). To address this issue, we propose a novel 1-D EEG signal denoising network with 2-D transformer, namely EEGDnet. Specifically, we comprehensively take into account the non-local and local self-similarity of EEG signal through the transformer module. By fusing non-local self-similarity in self-attention blocks and local self-similarity in feed forward blocks, the negative impact caused by noises and outliers can be reduced significantly. Extensive experiments show that, compared with other state-of-the-art models, EEGDnet achieves much better performance in terms of both quantitative and qualitative metrics. Specifically, EEGDnet can achieve 18% and 11% improvements in correlation coefficients when removing ocular artifacts and muscle artifacts, respectively.

Automatic Muscle Artifacts Identification and Removal from Single-Channel EEG Using Wavelet Transform with Meta-Heuristically Optimized Non-Local Means Filter

Electroencephalogram (EEG) signals may get easily contaminated by muscle artifacts, which may lead to wrong interpretation in the brain–computer interface (BCI) system as well as in various medical diagnoses. The main objective of this paper is to remove muscle artifacts without distorting the information contained in the EEG. A novel multi-stage EEG denoising method is proposed for the first time in which wavelet packet decomposition (WPD) is combined with a modified non-local means (NLM) algorithm. At first, the artifact EEG signal is identified through a pre-trained classifier. Next, the identified EEG signal is decomposed into wavelet coefficients and corrected through a modified NLM filter. Finally, the artifact-free EEG is reconstructed from corrected wavelet coefficients through inverse WPD. To optimize the filter parameters, two meta-heuristic algorithms are used in this paper for the first time. The proposed system is first validated on simulated EEG data and then tested on real EEG data. The proposed approach achieved average mutual information (MI) as 2.9684 ± 0.7045 on real EEG data. The result reveals that the proposed system outperforms recently developed denoising techniques with higher average MI, which indicates that the proposed approach is better in terms of quality of reconstruction and is fully automatic.

Automated ECG multi-class classification system based on combining deep learning features with HRV and ECG measures

Electrocardiogram (ECG) serves as the gold standard for noninvasive diagnosis of several types of heart disorders. In this study, a novel hybrid approach of deep neural network combined with linear and nonlinear features extracted from ECG and heart rate variability (HRV) is proposed for ECG multi-class classification. The proposed system enhances the ECG diagnosis performance by combining optimized deep learning features with an effective aggregation of ECG features and HRV measures using chaos theory and fragmentation analysis. The constant-Q non-stationary Gabor transform technique is employed to convert the 1-D ECG signal into 2-D image which is sent to a pre-trained convolutional neural network structure, called AlexNet. The pair-wise feature proximity algorithm is employed to select the optimal features from the AlexNet output feature vector to be concatenated with the ECG and HRV measures. The concatenated features are sent to different types of classifiers to distinguish three distinct subjects, namely congestive heart failure, arrhythmia, and normal sinus rhythm (NSR). The results reveal that the linear discriminant analysis classifier has the highest accuracy compared to the other classifiers. The proposed system is investigated with real ECG data taken from well-known databases, and the experimental results show that the proposed diagnosis system outperforms other recent state-of-the-art systems in terms of accuracy 98.75%, specificity 99.00%, sensitivity of 98.18%, and computational time 0.15 s. This demonstrates that the proposed system can be used to assist cardiologists in enhancing the accuracy of ECG diagnosis in real-time clinical setting.