Enhancing EEG and sEMG Fusion Decoding Using a Multi-Scale Parallel Convolutional Network With Attention Mechanism.

Electroencephalography (EEG) and surface electromyography (sEMG) have been widely used in the rehabilitation training of motor function. However, EEG signals have poor user adaptability and low classification accuracy in practical applications, and sEMG signals are susceptible to abnormalities such as muscle fatigue and weakness, resulting in reduced stability. To improve the accuracy and stability of interactive training recognition systems, we propose a novel approach called the Attention Mechanism-based Multi-Scale Parallel Convolutional Network (AM-PCNet) for recognizing and decoding fused EEG and sEMG signals. Firstly, we design an experimental scheme for the synchronous collection of EEG and sEMG signals and propose an ERP-WTC analysis method for channel screening of EEG signals. Then, the AM-PCNet network is designed to extract the time-domain, frequency-domain, and mixed-domain information of the EEG and sEMG fusion spectrogram images, and the attention mechanism is introduced to extract more fine-grained multi-scale feature information of the EEG and sEMG signals. Experiments on datasets obtained in the laboratory have shown that the average accuracy of EEG and sEMG fusion decoding is 96.62%. The accuracy is significantly improved compared with the classification performance of single-mode signals. When the muscle fatigue level reaches 50% and 90%, the accuracy is 92.84% and 85.29%, respectively. This study indicates that using this model to fuse EEG and sEMG signals can improve the accuracy and stability of hand rehabilitation training for patients.

Dynamic pruning group equivariant network for motor imagery EEG recognition.

Introduction: The decoding of the motor imaging electroencephalogram (MI-EEG) is the most critical part of the brain-computer interface (BCI) system. However, the inherent complexity of EEG signals makes it challenging to analyze and model them. Methods: In order to effectively extract and classify the features of EEG signals, a classification algorithm of motor imagery EEG signals based on dynamic pruning equal-variant group convolutional network is proposed. Group convolutional networks can learn powerful representations based on symmetric patterns, but they lack clear methods to learn meaningful relationships between them. The dynamic pruning equivariant group convolution proposed in this paper is used to enhance meaningful symmetric combinations and suppress unreasonable and misleading symmetric combinations. At the same time, a new dynamic pruning method is proposed to dynamically evaluate the importance of parameters, which can restore the pruned connections. Results and Discussion: The experimental results show that the pruning group equivariant convolution network is superior to the traditional benchmark method in the benchmark motor imagery EEG data set. This research can also be transferred to other research areas.

Motor Imagery EEG Decoding Based on Multi-scale Hybrid Networks and Feature Enhancement.

Motor Imagery (MI) based on Electroencephalography (EEG), a typical Brain-Computer Interface (BCI) paradigm, can communicate with external devices according to the brain's intentions. Convolutional Neural Networks (CNN) are gradually used for EEG classification tasks and have achieved satisfactory performance. However, most CNN-based methods employ a single convolution mode and a convolution kernel size, which cannot extract multi-scale advanced temporal and spatial features efficiently. What's more, they hinder the further improvement of the classification accuracy of MI-EEG signals. This paper proposes a novel Multi-Scale Hybrid Convolutional Neural Network (MSHCNN) for MI-EEG signal decoding to improve classification performance. The two-dimensional convolution is used to extract temporal and spatial features of EEG signals and the one-dimensional convolution is used to extract advanced temporal features of EEG signals. In addition, a channel coding method is proposed to improve the expression capacity of the spatiotemporal characteristics of EEG signals. We evaluate the performance of the proposed method on the dataset collected in the laboratory and BCI competition IV 2b, 2a, and the average accuracy is at 96.87%, 85.25%, and 84.86%, respectively. Compared with other advanced methods, our proposed method achieves higher classification accuracy. Then we use the proposed method for an online experiment and design an intelligent artificial limb control system. The proposed method effectively extracts EEG signals' advanced temporal and spatial features. Additionally, we design an online recognition system, which contributes to the further development of the BCI system.

Construction of a Competitive Endogenous RNA Network Related to Exosomes in Diabetic Retinopathy.

BACKGROUND: The competing endogenous RNA (ceRNA) network plays an important role in the occurrence and development of a variety of diseases. This study aimed to construct a ceRNA network related to exosomes in diabetic retinopathy (DR). METHODS: We explored the Gene Expression Omnibus (GEO) database and then analyzed the RNAs of samples to obtain differentially expressed lncRNAs (DELs), miRNAs (DEMs) and mRNAs (DEGs) alongside the progress of DR. Next, Gene Set Enrichment Analysis (GSEA) analysis of DEGs, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis of up-DEGs were performed. In addition, a ceRNA network related to exosomes in DR was constructed on the base of DELs, DEMs and DEGs. Finally, the function of the ceRNA network was explored by GO and KEGG enrichment analysis. RESULTS: Through our analysis, 267 DELs (93 up and 174 down), 114 DEMs (64 up and 50 down) and 2368 DEGs (1252 up and 1116 down) were screened. The GSEA analysis results show that these genes were mainly related to cytokine-cytokine receptor interaction, hippo signaling pathway and JAK-STAT signaling pathway. The GO and KEGG results show that these up-DEGs were mainly enriched in viral gene expression, components of ribosomes, mineral absorption, Wntprotein binding, and TGF-ß signaling pathway. Besides, a ceRNA network, including 15 lncRNAs (e.g., C1orf145, FGF14-IT1, and PRNT), 3 miRNAs (miR-10a-5p, miR-1297 and miR-507) and 11 mRNAs (NCOR2, CHAC1 and LIX1L, etc.) was constructed. Those 5 lncRNAs were up-regulated, 1 miRNA was down-regulated and 5 mRNAs were up-regulated in DR, while 10 lncRNAs were downregulated, 2 miRNAs were up-regulated and 6 mRNAs were down-regulated in DR. CONCLUSION: The novel ceRNA network that we constructed will provide new insights into the underlying molecular mechanisms of exosomes in DR.

Improved Stretchable and Sensitive Fe Nanowire-Based Strain Sensor by Optimizing Areal Density of Nanowire Network.

Flexible strain sensors, when considering high sensitivity and a large strain range, have become a key requirement for current robotic applications. However, it is still a thorny issue to take both factors into consideration at the same time. Here, we report a sandwich-structured strain sensor based on Fe nanowires (Fe NWs) that has a high GF (37-53) while taking into account a large strain range (15-57.5%), low hysteresis (2.45%), stability, and low cost with an areal density of Fe NWs of 4.4 mg/cm2. Additionally, the relationship between the contact point of the conductive network, the output resistance, and the areal density of the sensing unit is analyzed. Microscopically, the contact points of the conductive network directly affect the sensor output resistance distribution, thereby affecting the gauge factor (GF) of the sensor. Macroscopically, the areal density and the output resistivity of the strain sensor have the opposite percolation theory, which affects its linearity performance. At the same time, there is a positive correlation between the areal density and the contact point: when the stretching amount is constant, it theoretically shows that the areal density affects the GF. When the areal density reaches this percolation threshold range, the sensing performance is the best. This will lay the foundation for rapid applications in wearable robots.

FB-EEGNet: A fusion neural network across multi-stimulus for SSVEP target detection.

BACKGROUND: Steady-state visual evoked potential (SSVEP) is a prevalent paradigm of brain-computer interface (BCI). Recently, deep neural networks (DNNs) have been employed for SSVEP target recognition. However, current DNN models can not fully extract information from SSVEP harmonic components, and ignore the influence of non-target stimuli. NEW METHOD: To employ information of multiple sub-bands and non-target stimulus data, we propose a DNN model for SSVEP target detection, i.e., FB-EEGNet, which fuses features of multiple neural networks. Additionally, we design a multi-label for each sample and optimize the parameters of FB-EEGNet across multi-stimulus to incorporate the information from non-target stimuli. RESULTS: Under the subject-specific condition, FB-EEGNet achieves the average classification accuracies (information transfer rate (ITR)) of 76.75 % (50.70 bits/min) and 89.14 % (70.45 bits/min) in a time widow of 0.7 s under the public 12-target dataset and our experimental 9-target dataset, respectively. Under the cross-subject condition, FB-EEGNet achieved mean accuracies (ITRs) of 81.72 % (67.99 bits/min) and 92.15 % (76.12 bits/min) on the public and experimental datasets in a time window of 1 s, respectively. COMPARISON WITH EXISTING METHODS: FB-EEGNet shows superior performance than CCNN, EEGNet, CCA and FBCCA both for subject-dependent and subject-independent SSVEP target recognition. CONCLUSION: FB-EEGNet can effectively extract information from multiple sub-bands and cross-stimulus targets, providing a promising way for extracting deep features in SSVEP using neural networks.

Continuous Terminal Sliding-Mode Control for FJR Subject to Matched/Mismatched Disturbances.

A robust finite-time control (FTC) framework using continuous terminal sliding-mode control (SMC) and high-order sliding-mode observer (HOSMO) is discussed to realize the trajectory tracking of flexible-joint robots in this article. Control performances of the robots always suffer from unknown matched and mismatched time-varying disturbances. Traditional SMC exists with a chattering phenomenon and cannot cope with mismatched time-varying disturbances due to its inherent structure property. For this reason, two HOSMOs are devised to estimate the time-varying disturbances on the link and motor side, respectively. Then, by fusing the states and disturbance estimations into a novel terminal sliding-mode surface, a continuous robust FTC scheme is developed. The proposed control strategy can not only handle both matched and mismatched time-varying disturbances but also obtain a finite-time convergence performance. The rigorous finite-time stability analysis of the closed-loop system under the proposed control method is guaranteed. The results are illustrated to verify the effectiveness and robustness of the proposed design approach.

Multi-Focus Image Fusion Method Based on Multi-Scale Decomposition of Information Complementary.

Multi-focus image fusion is an important method used to combine the focused parts from source multi-focus images into a single full-focus image. Currently, to address the problem of multi-focus image fusion, the key is on how to accurately detect the focus regions, especially when the source images captured by cameras produce anisotropic blur and unregistration. This paper proposes a new multi-focus image fusion method based on the multi-scale decomposition of complementary information. Firstly, this method uses two groups of large-scale and small-scale decomposition schemes that are structurally complementary, to perform two-scale double-layer singular value decomposition of the image separately and obtain low-frequency and high-frequency components. Then, the low-frequency components are fused by a rule that integrates image local energy with edge energy. The high-frequency components are fused by the parameter-adaptive pulse-coupled neural network model (PA-PCNN), and according to the feature information contained in each decomposition layer of the high-frequency components, different detailed features are selected as the external stimulus input of the PA-PCNN. Finally, according to the two-scale decomposition of the source image that is structure complementary, and the fusion of high and low frequency components, two initial decision maps with complementary information are obtained. By refining the initial decision graph, the final fusion decision map is obtained to complete the image fusion. In addition, the proposed method is compared with 10 state-of-the-art approaches to verify its effectiveness. The experimental results show that the proposed method can more accurately distinguish the focused and non-focused areas in the case of image pre-registration and unregistration, and the subjective and objective evaluation indicators are slightly better than those of the existing methods.

Multi-Focus Color Image Fusion Based on Quaternion Multi-Scale Singular Value Decomposition.

Most existing multi-focus color image fusion methods based on multi-scale decomposition consider three color components separately during fusion, which leads to inherent color structures change, and causes tonal distortion and blur in the fusion results. In order to address these problems, a novel fusion algorithm based on the quaternion multi-scale singular value decomposition (QMSVD) is proposed in this paper. First, the multi-focus color images, which represented by quaternion, to be fused is decomposed by multichannel QMSVD, and the low-frequency sub-image represented by one channel and high-frequency sub-image represented by multiple channels are obtained. Second, the activity level and matching level are exploited in the focus decision mapping of the low-frequency sub-image fusion, with the former calculated by using local window energy and the latter measured by the color difference between color pixels expressed by a quaternion. Third, the fusion results of low-frequency coefficients are incorporated into the fusion of high-frequency sub-images, and a local contrast fusion rule based on the integration of high-frequency and low-frequency regions is proposed. Finally, the fused images are reconstructed employing inverse transform of the QMSVD. Simulation results show that image fusion using this method achieves great overall visual effects, with high resolution images, rich colors, and low information loss.

Introducing frequency representation into convolution neural networks for medical image segmentation via twin-Kernel Fourier convolution.

BACKGROUND AND OBJECTIVE: For medical image segmentation, deep learning-based methods have achieved state-of-the-art performance. However, the powerful spectral representation in the field of image processing is rarely considered in these models. METHODS: In this work, we propose to introduce frequency representation into convolution neural networks (CNNs) and design a novel model, tKFC-Net, to combine powerful feature representation in both frequency and spatial domains. Through the Fast Fourier Transform (FFT) operation, frequency representation is employed on pooling, upsampling, and convolution without any adjustments to the network architecture. Furthermore, we replace original convolution with twin-Kernel Fourier Convolution (t-KFC), a new designed convolution layer, to specify the convolution kernels for particular functions and extract features from different frequency components. RESULTS: We experimentally show that our method has an edge over other models in the task of medical image segmentation. Evaluated on four datasets-skin lesion segmentation (ISIC 2018), retinal blood vessel segmentation (DRIVE), lung segmentation (COVID-19-CT-Seg), and brain tumor segmentation (BraTS 2019), the proposed model achieves outstanding results: the metric F1-Score is 0.878 for ISIC 2018, 0.8185 for DRIVE, 0.9830 for COVID-19-CT-Seg, and 0.8457 for BraTS 2019. CONCLUSION: The introduction of spectral representation retains spectral features which result in more accurate segmentation. The proposed method is orthogonal to other topology improvement methods and very convenient to be combined.

Motor imagery EEG recognition with KNN-based smooth auto-encoder.

As new human-computer interaction technology, brain-computer interface has been widely used in various fields of life. The study of EEG signals cannot only improve people's awareness of the brain, but also establish new ways for the brain to communicate with the outside world. This paper takes the motion imaging EEG signal as the research object and proposes an innovative semi-supervised model called KNN-based smooth auto-encoder (k-SAE). K-SAE looks for the nearest neighbor values of the samples to construct a new input and learns the robust features representation by reconstructing this new input instead of the original input, which is different from the traditional automatic encoder (AE). The Gaussian filter is selected as the convolution kernel function in k-SAE to smooth the noise in the feature. Besides, the data information and spatial position of the feature map are recorded by max-pooling and unpooling, that help to prevent loss of important information. The method is applied to two data sets for feature extraction and classification experiments of motor imaging EEG signals. The experimental results show that k-SAE achieves good recognition accuracy and outperforms other state-of-the-art recognition algorithms.

Semisupervised Deep Stacking Network with Adaptive Learning Rate Strategy for Motor Imagery EEG Recognition.

Practical motor imagery electroencephalogram (EEG) data-based applications are limited by the waste of unlabeled samples in supervised learning and excessive time consumption in the pretraining period. A semisupervised deep stacking network with an adaptive learning rate strategy (SADSN) is proposed to solve the sample loss caused by supervised learning of EEG data and the extraction of manual features. The SADSN adopts the idea of an adaptive learning rate into a contrastive divergence (CD) algorithm to accelerate its convergence. Prior knowledge is introduced into the intermediary layer of the deep stacking network, and a restricted Boltzmann machine is trained by a semisupervised method in which the adjusting scope of the coefficient in learning rate is determined by performance analysis. Several EEG data sets are carried out to evaluate the performance of the proposed method. The results show that the recognition accuracy of SADSN is advanced with a more significant convergence rate and successfully classifies motor imagery.