A simple and effective deep neural network based QRS complex detection method on ECG signal.

Introduction: The QRS complex is the most prominent waveform within the electrocardiograph (ECG) signal. The accurate detection of the QRS complex is an essential step in the ECG analysis algorithm, which can provide fundamental information for the monitoring and diagnosis of the cardiovascular diseases. Methods: Seven public ECG datasets were used in the experiments. A simple and effective QRS complex detection algorithm based on the deep neural network (DNN) was proposed. The DNN model was composed of two parts: a feature pyramid network (FPN) based backbone with dual input channels to generate the feature maps, and a location head to predict the probability of point belonging to the QRS complex. The depthwise convolution was applied to reduce the parameters of the DNN model. Furthermore, a novel training strategy was developed. The target of the DNN model was generated by using the points within 75 milliseconds and beyond 150 milliseconds from the closest annotated QRS complexes, and artificial simulated ECG segments with high heart rates were generated in the data augmentation. The number of parameters and floating point operations (FLOPs) of our model was 26976 and 9.90M, respectively. Results: The proposed method was evaluated through a cross-dataset test and compared with the sophisticated state-of-the-art methods. On the MITBIH NST, the proposed method demonstrated slightly better sensitivity (95.59% vs. 95.55%) and lower presicion (91.03% vs. 92.93%). On the CPSC 2019, the proposed method have similar sensitivity (95.15% vs.95.13%) and better precision (91.75% vs. 82.03%). Discussion: Experimental results show the proposed algorithm achieved a comparable performance with only a few parameters and FLOPs, which would be useful for the application of ECG analysis on the wearable device.

LDLR is an entry receptor for Crimean-Congo hemorrhagic fever virus.

Crimean-Congo hemorrhagic fever virus (CCHFV) is the most widespread tick-born zoonotic bunyavirus that causes severe hemorrhagic fever and death in humans. CCHFV enters the cell via clathrin-mediated endocytosis which is dependent on its surface glycoproteins. However, the cellular receptors that are required for CCHFV entry are unknown. Here we show that the low density lipoprotein receptor (LDLR) is an entry receptor for CCHFV. Genetic knockout of LDLR impairs viral infection in various CCHFV-susceptible human, monkey and mouse cells, which is restored upon reconstitution with ectopically-expressed LDLR. Mutagenesis studies indicate that the ligand binding domain (LBD) of LDLR is necessary for CCHFV infection. LDLR binds directly to CCHFV glycoprotein Gc with high affinity, which supports virus attachment and internalization into host cells. Consistently, a soluble sLDLR-Fc fusion protein or anti-LDLR blocking antibodies impair CCHFV infection into various susceptible cells. Furthermore, genetic knockout of LDLR or administration of an LDLR blocking antibody significantly reduces viral loads, pathological effects and death following CCHFV infection in mice. Our findings suggest that LDLR is an entry receptor for CCHFV and pharmacological targeting of LDLR may provide a strategy to prevent and treat Crimean-Congo hemorrhagic fever.

High expression ITGA2 affects the expression of MET, PD-L1, CD4 and CD8 with the immune microenvironment in pancreatic cancer patients.

Purpose: Pancreatic cancer is characterized by a grim prognosis and is regarded as one of the most formidable malignancies. Among the genes exhibiting high expression in different tumor tissues, ITGA2 stands out as a promising candidate for cancer therapy. The promotion of cancer in pancreatic cancer is not effective. The objective of this study is to assess the presence of ITGA2, EMT and PD-L1 in pancreatic cancer. Experimental design: We examined the expression of ITGA2, MET, E-cadherin, PD-L1, CD4, and CD8 proteins in 62 pancreatic cancer tissue samples using multi-tissue immunofluorescence and immunohistochemistry techniques. Functional assays, such as the cell migration assay and transwell assay, were used to determine the biological role of ITGA2 in pancreatic cancer. The relationship of ITGA2,EMT and PD-L1 were examined using Western blot analysis and RT-qPCR assay. Results: In our study, we observed the expression of ITGA2, E-cadherin, and PD-L1 in both tumor and stroma tissues of pancreatic cancer. Additionally, a positive correlation between ITGA2, E-cadherin, and PD-L1 in the tumor region (r=0.559, P<0.001 and r=0.511, P<0.001), and PD-L1 in the stroma region (r=0.512, P<0.001).The expression levels of ITGA2, CD4, and CD8 were found to be higher in pancreatic cancer tissues compared to adjacent tissues (P < 0.05). Additionally, ITGA2 was negatively correlated with CD4 and CD8 (r = -0.344, P < 0.005 and r = -0.398, P < 0.005).Furthermore, ITGA2, CD4, and CD8 were found to be correlated with the survival time of patients (P < 0.05). Blocking ITGA2 inhibited the proliferation and invasion ability of pancreatic cancer cells significantly, Additionally, sh-ITGA2 can down-regulate the expression of EMT and PD-L1. Conclusions: We identified a novel mechanism in which ITGA2 plays a crucial role in the regulation of pancreatic cancer growth and invasion. This mechanism involves the upregulation of MET and PD-L1 expression in pancreatic cancer cells. Additionally, we found that increased expression of ITGA2 is associated with a poor prognosis in pancreatic cancer patients. Furthermore, ITGA2 also affects immune regulation in these patients. Therefore, targeting ITGA2 is an effective method to enhance the efficacy of checkpoint immunotherapy and prohibiting tumor growth against pancreatic cancer.

Body mass index, C-reactive protein, and pancreatic cancer: A Mendelian randomization analysis to investigate causal pathways.

Aim: To explore whether C-reactive protein (CRP) mediates the risk of body mass index (BMI) in pancreatic cancer (PC) and calculate the mediate proportion of CRP in this possible mechanism. Methods: Based on two-sample Mendelian randomization (TSMR), a two-step Mendelian randomization (TM) model was conducted to determine whether CRP was a mediator of the causal relationship between BMI and PC. The multivariable Mendelian randomization (MVMR) study was designed for mediating analysis and to calculate the mediating proportion mediated by CRP. Results: BMI has a positive causal relationship with PC (n = 393 SNPs, OR = 1.484, 95% CI: 1.021-2.157, p< 0.05). BMI has a positive causal relationship with CRP (n = 179 SNPs, OR = 1.393, 95% CI: 1.320-1.469, p< 0.05). CRP has a positive causal relationship with PC (n = 54 SNPs, OR = 1.348, 95% CI: 1.004-1.809, p< 0.05). After adjusting CRP, BMI has no causal relationship with PC (n = 334 SNPs, OR = 1.341, 95% CI: 0.884-2.037, p< 0.05). After adjusting BMI, there was still a positive causal relationship between CRP and PC (n = 334 SNPs, OR = 1.441, 95% CI: 1.064-1.950, p< 0.05). The mediating effect of CRP was 29%. Conclusions: In clinical practice, while actively advocating for weight loss among obese patients, we should focus on chronic inflammation levels in obese patients as well. In addition, anti-inflammatory dietary patterns and appropriate physical activity are important in preventing PC.

Assessing the causality between thyroid and breast neoplasms: A bidirectional Mendelian randomization study.

Aim: This study aimed to evaluate the association between thyroid neoplasms (TN) and the risk of developing breast neoplasms (BN) by assessing data on single nucleotide polymorphisms (SNPs) obtained from the Deutsches Krebsforschungszentrum (DKFZ) and Breast Cancer Association (BCAC). Methods: Data on SNPs associated with TN and BN were obtained from DKFZ and BCAC, respectively. Secondary data analysis of all pooled data from genome-wide association studies (GWAS) was performed to identify the genetic loci closely associated with TN or BN as instrumental variables (IVs). To evaluate the causal relationship between TN and BN, a bidirectional Mendelian randomization (MR) analysis was performed using MR Egger regression, weighted median, inverse variance weighted (IVW) random effects model, simple mode, weighted mode, maximum likelihood, penalized weighted median, IVW radial, IVW fixed effects, and robust adjusted profile scores (RAPS) method. Results: The MR in this study demonstrated a modest reverse causal relationship between TN and BN but a significant positive causal relationship between BN and TN. Conclusions: The MR of this study provided genetic evidence suggesting an association between BN and TN; however, further research is warranted to explore the potential mechanism of interaction between these two malignancies. Moreover, general breast screening should be performed in individuals with TN, but TN screening should be reinforced in individuals with BN.

Mex3B inhibits DC-STAMP mRNA level and osteoclastogenesis.

Bone homeostasis is maintained through continuous remodeling by osteoclast-driven bone resorption and osteoblast-mediated bone formation. Osteoclasts are multinucleated giant cells (MNCs) differentiated from myeloid progenitors of the monocytic lineage. During osteoclast maturation, DC-STAMP (dendritic cell specific transmembrane protein) has been shown as a master determinant of osteoclast cell fusion. In this study, we demonstrate that Mex3B inhibits osteoclast fusion protein DCSTAMP expression and osteoclastogenesis. During differentiation of osteoclasts, the expression of Mex3B is down-regulated by cytokines such as RANKL and TNFa, resulting in relief of Mex3B-mediated down-regulation of DC-STAMP mRNA level. Our findings not only reveal critical mechanisms on regulation of DC-STAMP-mediated osteoclastogenesis, but also point to Mex3B as a potential therapeutic target for the treatment of human bone diseases.

Deep Multi-instance Networks for Bundle Branch Block Detection from Multi-lead ECG.

Bundle branch block (BBB) is one of the most common cardiac disorder, and can be detected by electro-cardiogram (ECG) signal in clinical practice. Conventional methods adopted some kinds of hand-craft features, whose discriminative power is relatively low. On the other hand, these methods were based on the supervised learning, which required the high cost heartbeat annotation in the training. In this paper, a novel end-to-end deep network was proposed to classify three types of heartbeat: right BBB (RBBB), left BBB (LBBB) and others with a multiple instance learning based training strategy. We trained the proposed method on the China Physiological Signal Challenge 2018 database (CPSC) and tested on the MIT-BIH Arrhythmia database (AR). The proposed method achieved an accuracy of 78.58%, and sensitivity of 84.78% (LBBB), 51.23% (others) and 99.72% (RBBB), better than the baseline methods. Experimental results show that our method would be a good choice for the BBB classification on the ECG dataset with record-level labels instead of heartbeat annotations.

A Weighted Graph Attention Network Based Method for Multi-label Classification of Electrocardiogram Abnormalities.

The multi-label electrocardiogram (ECG) classification is to automatically predict a set of concurrent cardiac abnormalities in an ECG record, which is significant for clinical diagnosis. Modeling the cardiac abnormality dependencies is the key to improving classification performance. To capture the dependencies, we proposed a multi-label classification method based on the weighted graph attention networks. In the study, a graph taking each class as a node was mapped and the class dependencies were represented by the weights of graph edges. A novel weights generation method was proposed by combining the self-attentional weights and the prior learned co-occurrence knowledge of classes. The algorithm was evaluated on the dataset of the Hefei Hi-tech Cup ECG Intelligent Competition for 34 kinds of ECG abnormalities classification. And the micro-f 1 and the macro-f 1 of cross validation respectively were 91.45% and 44.48%. The experiment results show that the proposed method can model class dependencies and improve classification performance.

SAE+LSTM: A New Framework for Emotion Recognition From Multi-Channel EEG.

EEG-based automatic emotion recognition can help brain-inspired robots in improving their interactions with humans. This paper presents a novel framework for emotion recognition using multi-channel electroencephalogram (EEG). The framework consists of a linear EEG mixing model and an emotion timing model. Our proposed framework considerably decomposes the EEG source signals from the collected EEG signals and improves classification accuracy by using the context correlations of the EEG feature sequences. Specially, Stack AutoEncoder (SAE) is used to build and solve the linear EEG mixing model and the emotion timing model is based on the Long Short-Term Memory Recurrent Neural Network (LSTM-RNN). The framework was implemented on the DEAP dataset for an emotion recognition experiment, where the mean accuracy of emotion recognition achieved 81.10% in valence and 74.38% in arousal, and the effectiveness of our framework was verified. Our framework exhibited a better performance in emotion recognition using multi-channel EEG than the compared conventional approaches in the experiments.

A Novel Detection Method of Bundle Branch Block from Multi-lead ECG.

Bundle branch block (BBB) is a common conduction block disease and can be diagnosed using electrocardiogram (ECG) signal in clinical practice. In this paper, a novel method was proposed to detect two types of BBB: right BBB (RBBB) and left BBB (LBBB) based on the combination of deep features and several kinds of expert features. We evaluated the proposed method on the MIT-BIH Arrhythmia database (AR) and China Physiological Signal Challenge 2018 database (CPSC). The proposed method achieved an accuracy of 99.96% (AR) in the class-oriented evaluation and an accuracy of 98.76% (AR) and 97.88% (CPSC) in the subject-oriented evaluation, better than the baseline methods. Experimental results show that our method would be a good choice for the detection of the BBB.

Deep Learning Based Patient-Specific Classification of Arrhythmia on ECG signal.

The classification of the heartbeat type is an essential function in the automatical electrocardiogram (ECG) analysis algorithm. The guideline of the ANSI/AAMI EC57 defined five types of heartbeat: non-ectopic or paced beat (N), supraventricular ectopic beat (S), ventricular ectopic beat (V), fusion of a ventricular and normal beat (F), pace beat or fusion of a paced and a normal or beat that cannot be classified (Q). In the work, a deep neural network based method was proposed to classify these five types of heartbeat. After removing the noise from ECG signals by a low-pass filter, the two-lead heartbeat segments with 2-s length were generated on the filtered signals, and classified by an adaptive ResNet model. The proposed method was evaluated on the MIT-BIH Arrhythmia Database with the patient-specific pattern. The overall accuracy was 98.6% and sensitivity of N, S, V, F were 99.4%, 85.4%, 96.6%, 90.6% respectively. Experimental results show that the proposed method achieved a good performance, and would be useful in the clinic practice.

An Electrocardiogram Delineator via Deep Segmentation Network.

Electrocardiogram (ECG) delineation is a process to detect multiple characteristic points, which contain critical diagnostic information about cardiac diseases. We treat the ECG delineation task as an one-dimensional segmentation problem, and propose a novel end-to-end deep learning method to segment sections of ECG signal. Our neural network consists of two parts: a segmentation network composed of multiple 1D Convolutional Neural Networks (CNN) and a postprocessing network composed of a sequential Conditional Random Field (CRF). Our method is trained and validated on QT database. The experimental results show that our method yields competitive overall performance compared with other state-of-the-art works and outperform them on onset of the P wave and offset of the T wave.

Novel Deep Convolutional Neural Network for Cuff-less Blood Pressure Measurement Using ECG and PPG Signals.

Cuff-less blood pressure (BP) is a potential method for BP monitoring because it is undisturbed and continuous monitoring. Existing cuff-less estimation methods are easily influenced by signal noise and non-ideal signal morphology. In this study we propose a novel well-designed Convolutional Neural Network (CNN) model named Deep-BP for BP estimation task. The structure of Deep-BP can help to capture more underlying data features associated with BP than handcrafted features, thus improving the robustness and estimation accuracy. We carry out experiments with and without calibration procedure in training stage to evaluate the performance of new method in different application scenarios. The experiment results show that the Deep-BP model achieves high accuracy and outperforms existing methods, in the experiments both with and without calibration.

A Novel Method for Calibration-Based Cuff-Less Blood Pressure Estimation.

Cuff-less blood pressure estimation technology is useful for cardiovascular disease monitoring. However, without calibration, cuff-less blood pressure estimation is hard to achieve clinical acceptable performance. The traditional methods are always calibrated with retraining. With the increases of the parameters number, the cost of model retraining increases several times. So we propose a novel blood pressure estimation method, which can be calibrated with reference inputs rather than with retraining. The experiment results suggest that the method we proposed can achieve clinical performance (SBP:-0.004 ± 5.869 mmHg, DBP:-0.004±4.511 mmHg) with low calibration cost.

A Review of Emotion Recognition Using Physiological Signals.

Emotion recognition based on physiological signals has been a hot topic and applied in many areas such as safe driving, health care and social security. In this paper, we present a comprehensive review on physiological signal-based emotion recognition, including emotion models, emotion elicitation methods, the published emotional physiological datasets, features, classifiers, and the whole framework for emotion recognition based on the physiological signals. A summary and comparation among the recent studies has been conducted, which reveals the current existing problems and the future work has been discussed.

Overexpression of Malonyl-CoA: ACP Transacylase in Schizochytrium sp. to Improve Polyunsaturated Fatty Acid Production.

Polyunsaturated fatty acids (PUFAs) have been widely applied in the food and medical industry. In this study, malonyl-CoA: ACP transacylase (MAT) was overexpressed through homologous recombination to improve PUFA production in Schizochytrium. The results showed that the lipid and PUFA concentration were increased by 10.1 and 24.5% with MAT overexpression, respectively. Metabolomics analysis revealed that the intracellular tricarboxylic acid cycle was weakened and glucose absorption was accelerated in the engineered strain. In the mevalonate pathway, intracellular carotene content was decreased, and the carbon flux was then redirected toward PUFA synthesis. Furthermore, a glucose fed-batch fermentation was finally performed with the engineered Schizochytrium. The total lipid yield was further increased to 110.5 g/L, 39.6% higher than the wild strain. Docosahexaenoic acid and eicosapentaenoic acid yield were enhanced to 47.39 g/L and 1.65 g/L with an increase of 81.5 and 172.5%, respectively. This study provided an effective metabolic engineering strategy for industrial PUFA production.

CNVD: text mining-based copy number variation in disease database.

Copy number variation (CNV) is a kind of chromosomal structural reorganization that has been detected, in this decade, mainly by high-throughput biological technology. Researchers have found that CNVs are ubiquitous in many species and accumulating evidence indicates that CNVs are closely related with complex diseases. The investigation of chromosomal structural alterations has begun to reveal some important clues to the pathologic causes of diseases and to the disease process. However, many of the published studies have focused on a single disease and, so far, the experimental results have not been systematically collected or organized. Manual text mining from 6301 published papers was used to build the Copy Number Variation in Disease database (CNVD). CNVD contains CNV information for 792 diseases in 22 species from diverse types of experiments, thus, ensuring high confidence and comprehensive representation of the relationship between the CNVs and the diseases. In addition, multiple query modes and visualized results are provided in the CNVD database. With its user-friendly interface and the integrated CNV information for different diseases, CNVD will offer a truly comprehensive platform for disease research based on chromosomal structural variations. The CNVD interface is accessible at http://bioinfo.hrbmu.edu.cn/CNVD.

Calculating phenotypic similarity between genes using hierarchical structure data based on semantic similarity.

Phenotypic similarity is correlated with a number of measures of gene function, such as relatedness at the level of direct protein-protein interaction. The phenotypic effect of a deleted or mutated gene, which is one part of gene annotation, has caught broad attention. However, there have been few measures to study phenotypic similarity with the data from Human Phenotype Ontology (HPO) database, therefore more analogous measures should be developed and investigated. We used five semantic similarity-based measures (Jiang and Conrath, Lin, Schlicker, Yu and Wu) to calculate the human phenotypic similarity between genes (PSG) with data from HPO database, and evaluated their accuracy with information of protein-protein interaction, protein complex, protein family, gene function or DNA sequence. Compared with the gene pairs that were random selected, the results of these methods were statistically significant (all P<0.001). Furthermore, we assessed the performance of these five measures by receiver operating characteristic (ROC) curve analysis, and found that most of them performed better than the previous methods. This work had proved that these measures based on semantic similarity for calculation of PSG were effective for hierarchical structure data. Our study contributes to the development and optimization of novel algorithms of PSG calculation and provides more alternative methods to researchers as well as tools and directions for PSG study.

The application of gene co-expression network reconstruction based on CNVs and gene expression microarray data in breast cancer.

Copy number variations (CNVs) are one type of the human genetic variations and are pervasive in the human genome. It has been confirmed that they can play a causal role in complex diseases. Previous studies of CNVs focused more on identifying the disease-specific CNV regions or candidate genes on these CNV regions, but less on the synergistic actions between genes on CNV regions and other genes. Our research combined the CNVs with related gene co-expression to reconstruct gene co-expression network by using single nucleotide polymorphism microarray datasets and gene microarray datasets of breast cancer, and then extracted the modules which connected densely inside and analyzed the functions of modules. Interestingly, all of these modules' functions were related to breast cancer according to our enrichment analysis, and most of the genes in these modules have been reported to be involved in breast cancer. Our findings suggested that integrating CNVs and gene co-expressed relations was an available way to analyze the roles of CNV genes and their synergistic genes in breast cancer, and provided a novel insight into the pathological mechanism of breast cancer.

Analysis of functional and pathway association of differential co-expressed genes: a case study in drug addiction.

Drug addiction has been considered as a kind of chronic relapsing brain disease influenced by both genetic and environmental factors. At present, many causative genes and pathways related to diverse kinds of drug addiction have been discovered, while less attention has been paid to common mechanisms shared by different drugs underlying addiction. By applying a co-expression meta-analysis method to mRNA expression profiles of alcohol, cocaine, heroin addicted and normal samples, we identified significant gene co-expression pairs. As co-expression networks of drug group and control group constructed, associated function term pairs and pathway pairs reflected by co-expression pattern changes were discovered by integrating functional and pathway information respectively. The results indicated that respiratory electron transport chain, synaptic transmission, mitochondrial electron transport, signal transduction, locomotory behavior, response to amphetamine, negative regulation of cell migration, glucose regulation of insulin secretion, signaling by NGF, diabetes pathways, integration of energy metabolism, dopamine receptors may play an important role in drug addiction. In addition, the results can provide theory support for studies of addiction mechanisms.