A U-Net incorporating improved Transformer and convolutional channel attention module is designed for biventricular segmentation in MRI image.By replacing the high-level convolution of U-Net with the improved Transformer,the global feature information can be effectively extracted to cope with the challenge of poor segmentation performance due to the complex morphological variation of the right ventricle.The improved Transformer incorporates a fixed window attention for position localization in the self-attention module,and aggregates the output feature map for reducing the feature map size;and the network learning capability is improved by increasing network depth through the adjustment of multilayer perceptron.To solve the problem of unsatisfactory segmentation performance caused by blurred tissue edges,a feature aggregation module is used for the fusion of multi-level underlying features,and a convolutional channel attention module is adopted to rescale the underlying features to achieve adaptive learning of feature weights.In addition,a plug-and-play feature enhancement module is integrated to improve the segmentation performance which is affected by feature loss due to channel decay in the codec structure,which guarantees the spatial information while increasing the proportion of useful channel information.The test on the ACDC dataset shows that the proposed method has higher biventricular segmentation accuracy,especially for the right ventricle segmentation.Compared with other methods,the proposed method improves the DSC coefficient by at least 2.83%,proving its effectiveness in biventricular segmentation.

Objective To develop a multi-modal deep learning method for automatic classification of immune-mediated glomerular diseases based on images of optical microscopy(OM),immunofluorescence microscopy(IM),and transmission electron microscopy(TEM).Methods We retrospectively collected the pathological images from 273 patients and constructed a multi-modal multi-instance model for classification of 3 immune-mediated glomerular diseases,namely immunoglobulin A nephropathy(IgAN),membranous nephropathy(MN),and lupus nephritis(LN).This model adopts an instance-level multi-instance learning(I-MIL)method to select the TEM images for multi-modal feature fusion with the OM images and IM images of the same patient.By comparing this model with unimodal and bimodal models,we explored different combinations of the 3 modalities and the optimal methods for modal feature fusion.Results The multi-modal multi-instance model combining OM,IM,and TEM images had a disease classification accuracy of(88.34±2.12)%,superior to that of the optimal unimodal model[(87.08±4.25)%]and that of the optimal bimodal model[(87.92±3.06)%].Conclusion This multi-modal multi-instance model based on OM,IM,and TEM images can achieve automatic classification of immune-mediated glomerular diseases with a good classification accuracy.

Objective To develop a multi-modal deep learning method for automatic classification of immune-mediated glomerular diseases based on images of optical microscopy(OM),immunofluorescence microscopy(IM),and transmission electron microscopy(TEM).Methods We retrospectively collected the pathological images from 273 patients and constructed a multi-modal multi-instance model for classification of 3 immune-mediated glomerular diseases,namely immunoglobulin A nephropathy(IgAN),membranous nephropathy(MN),and lupus nephritis(LN).This model adopts an instance-level multi-instance learning(I-MIL)method to select the TEM images for multi-modal feature fusion with the OM images and IM images of the same patient.By comparing this model with unimodal and bimodal models,we explored different combinations of the 3 modalities and the optimal methods for modal feature fusion.Results The multi-modal multi-instance model combining OM,IM,and TEM images had a disease classification accuracy of(88.34±2.12)%,superior to that of the optimal unimodal model[(87.08±4.25)%]and that of the optimal bimodal model[(87.92±3.06)%].Conclusion This multi-modal multi-instance model based on OM,IM,and TEM images can achieve automatic classification of immune-mediated glomerular diseases with a good classification accuracy.

OBJECTIVE: We propose a novel region- level self-supervised contrastive learning method USRegCon (ultrastructural region contrast) based on the semantic similarity of ultrastructures to improve the performance of the model for glomerular ultrastructure segmentation on electron microscope images. METHODS: USRegCon used a large amount of unlabeled data for pre- training of the model in 3 steps: (1) The model encoded and decoded the ultrastructural information in the image and adaptively divided the image into multiple regions based on the semantic similarity of the ultrastructures; (2) Based on the divided regions, the first-order grayscale region representations and deep semantic region representations of each region were extracted by region pooling operation; (3) For the first-order grayscale region representations, a grayscale loss function was proposed to minimize the grayscale difference within regions and maximize the difference between regions. For deep semantic region representations, a semantic loss function was introduced to maximize the similarity of positive region pairs and the difference of negative region pairs in the representation space. These two loss functions were jointly used for pre-training of the model. RESULTS: In the segmentation task for 3 ultrastructures of the glomerular filtration barrier based on the private dataset GlomEM, USRegCon achieved promising segmentation results for basement membrane, endothelial cells, and podocytes, with Dice coefficients of (85.69 ± 0.13)%, (74.59 ± 0.13)%, and (78.57 ± 0.16)%, respectively, demonstrating a good performance of the model superior to many existing image-level, pixel-level, and region-level self-supervised contrastive learning methods and close to the fully- supervised pre-training method based on the large- scale labeled dataset ImageNet. CONCLUSION USRegCon facilitates the model to learn beneficial region representations from large amounts of unlabeled data to overcome the scarcity of labeled data and improves the deep model performance for glomerular ultrastructure recognition and boundary segmentation.
Humans ; Electrons ; Endothelial Cells ; Learning ; Podocytes ; Kidney Diseases

Methods:Cultured human alveolar epithelial A549 cells were assigned into LPS group and blank control group. LPS group was stimulated with LPS and adenosine triphosphate to induce pyroptosis and inflammation. A549 cells were divided into 4 groups: miR-20a mimics group, mimics-negative control (NC) group, inhibitor group and inhibitor-NC group. MiRNA-20a mimics, mimics-NC, inhibitor, and inhibitor-NC were transfected respectively into A549 cells, and after 24 h, the cells were collected to verify transfection efficiency by qPCR. MiRNA-20a mimics and the constructed TLR4-3'UTR double luciferase reporter plasmid were co-transfected into A549 cells, and luciferase activity was analyzed. MiRNA-20a mimics/inhibitors were transfected into A549 cells, and then the cells were stimulated by LPS for 8 h followed by adenosine triphosphate for 30 min. QPCR, Western Blot and ELISA were used to detect the expression of GSDMD, inflammatory factors (ASC, NLRP3, Caspase-1, IL-1β) and Signaling molecules (TLR4、NF-κB) in A549 cells at mRNA level and protein level. Immunofluorescence was used to detect the expression of TLR4 in the A549 cells and NF-κB in the nucleus of A549 cells after transfecting with miRNA-20a mimics/inhibitor.Results:The mRNA and protein expression of pyroptosis marker molecule (GSDMD) and inflammatory factors (ASC, NLRP3, Caspase-1, IL-1β) in A549 cells stimulated with LPS were significantly higher than those in the blank control group, and the differences were statistically significant ( P<0.05). The expression of miRNA-20 in the mimics group was significantly higher than that in the mimic-NC group ( P<0.05), while the expression of miRNA-20a in the inhibitor group was lower than that in the inhibitor-NC group ( P<0.01). The double luciferase reporter gene experiment showed that the relative fluorescence value of the co-transfection group for TLR4-3'UTR-WT and miRNA-20a mimics was significantly lower than the co-transfection group for TLR4-3'UTR-WT and miRNA-20a mimics-NC ( P<0.05). The mRNA and protein levels of pyroptosis marker molecule (GSDMD) , inflammatory factors (ASC, NLRP3, Caspase-1, IL-1β) and signaling molecules (TLR4, NF-κB) were decreased in the mimics group compared to the mimics-NC group, and increased in inhibitor group compared to inhibitor-NC group. Conclusions:miRNA-20a may inhibit LPS-induced pyroptosis and inflammation of A549 cells via TLR4/NF-κB signal pathway.Objetive:To explore the potential role of miRNA-20a in lipopolysaccharide (LPS) induced pyroptosis and inflamation of human alveolar epithelial A549 cells and its regulation mechanisim.

Allografts ; COVID-19 ; China/epidemiology* ; Disease Outbreaks ; Humans ; Kidney Transplantation/adverse effects* ; SARS-CoV-2

Objective:Differentially expressed circ_Dock6 was screened in vivo by applying circRNA high-throughput sequencing technology in lung tissue of newborn rats suffering from acute respiratory distress syndrome (ARDS). The corresponding target genes of microRNAs were predicted by bioinformatics, and their biological processes and signal pathways were analyzed as well. Methods:Real-time quantitative PCR was utilized to detect the expression of circ_Dock6 in the lung tissue of newborn rats in ARDS group (12 cases) and normal control group (12 cases). TargetScan, RNAhybrid and miRanda databases were adopted to predict the possible recruitment of miRNAs and their corresponding target genes by circ_Dock6.Functional enrichment analysis and signal pathway enrichment analysis were carried out on the target genes of each miRNA.Results:The expression of circ_Dock6 (0.44±0.29) in the lung tissue of ARDS group was significantly down-regulated ( t=2.060, P<0.05) compared with normal control group(1.63±1.33). The target gene intersections of miRNAs (miR-24-3p, miR-667-3p, miR-711, miR-203b-5p, miR-5132-5p, etc.) may be recruited by circ_Dock6 and were obtained from three databases.Its target gene aggregation function was enriched in various biological processes, including protein metabolism, protein amino acid phosphorylation, DNA-dependent transcriptional regulation, biological regulation, tissue and organ development, cell differentiation, signal regulation, gene expression, response to stimuli, almost all cellular components such as intracellular, organelle, cytoplasm, and nucleus, as well as molecular functions such as transferase activity, transcription factor activity, and phosphotransferase activity.The involved signaling pathways, including enrichment in mitogen-activated protein kinase(MAPK) signaling pathway, phosphatidylinositol-3-kinase-protein kinase B(PI3K-Akt)signaling pathway, and mammalian rapamycin target protein(mTOR)signaling pathway, were closely related to ARDS.Circ_Dock6 may play a significant role in the pathogenesis of ARDS. Conclusions:Circ_Dock6 may be closely correlated with the pathogenesis of neonatal ARDS.Through bioinformatics analysis, the prediction of its target genes and related signaling pathways laid the foundation for further explorations of its mechanism of action.

OBJECTIVE: To investigate the functional pathways enriched and differentially expressed genes (DEGs) in peripheral blood mononuclear cells (PBMCs) of patients with gram-positive and gram-negative sepsis. METHODS: Dataset GSE9960 obtained from NCBI GEO database containing PBMC samples from 16 non-infectious systematic inflammatory response syndrome (SIRS) patients, 17 gram-positive septic patients and 18 gram-negative septic patients were included in the study. Functional pathway annotations were conducted by gene set enrichment analysis and weighted gene co-expression network analysis. DEGs were filtered and master DEGs were then validated in PBMCs of gram-positive septic, gram-negative septic and non-infectious SIRS patients. RESULTS: The enriched gene sets in gram-positive sepsis and gram-negative sepsis were significantly different. The results indicated the opposite co-expression networks in SIRS and gram-negative sepsis, and the entirely different co-expression networks in gram-positive and gram-negative sepsis. Furthermore, we validated that CONCLUSIONS The results indicate that there are differences in the mechanism and pathogenesis of gram-positive and gram-negative sepsis, which may provide potential markers for sepsis diagnosis and empirical antimicrobial therapy.
Biomarkers/analysis* ; Gene Expression Profiling ; Gram-Negative Bacterial Infections/physiopathology* ; Gram-Positive Bacterial Infections/physiopathology* ; Humans ; Leukocytes, Mononuclear/pathology* ; Sepsis/physiopathology*

Objective To study the effect of small interfering ribonucleic acid (siRNA) silencing apoptosis signal-regulating kinase 1 (ASK1) on inflammatory response of lipopolysaccharide-induced alveolar epithelial A549 cells and its mechanism.Method Cell inflammation model of A549 cells was induced by lipopolysaccharide.The expression of ASK 1 in A549 cells was silenced by liposome transfection of siRNA.The mRNA and expression levels of ASK1,interleukin 6 (IL-6),interleukin 8 (IL-8) and tumor necrosis factor alpha (TNF-α) in A549 cells were detected by immunoblotting,real-time fluorescence quantitative polymerase chain reaction and enzyme-linked immunosorbent assay.Result The expression of IL-6,IL-8 and TNF-α in the experimental group was significantly higher than that in the control group (P＜0.001),which indicated that the inflammatory model of A549 cells was successfully constructed.The mRNA level and expression of ASK1 in the interference group was significantly lower than that in the negative control group and the blank control group (P＜0.01),indicating that silencing ASK1 was also successful.The expressions of IL-6,IL-8 and TNF-α in the interference group (0.37±0.04,0.32±0.04,0.48 ±0.13) were significantly lower than those in the negative control group (1.04±0.11,1.22±0.19,0.93±0.14) and the blank control group (1.01±0.14,1.01 ±0.23,1.02±0.25).The expression of IL-6,IL-8 and TNF-α protein in the interference group (pg/ml) (122.6± 11.0,537.2±42.4,159.2± 19.6) were also significantly lower than those in the negative control group (267.4±20.4,1 289.8±55.3,327.0±26.3) and blank control group (246.6±18.7,1 300.3±35.6,325.2± 18.3),with significant difference (P＜0.05).There was no significant difference in each value between negative control group and blank control group (P＞0.05).Conclusion Silencing ASK1 by siRNA can down-regulate the expression of IL-6,IL-8 and TNF-α in A549 cells,suggesting that ASK 1 may be involved in the regulation of lipopolysaccharide-induced inflammation in A549 cells.

Artificial joint replacement is an important means for the treatment of severe joint end-stage diseases such as hip and knee joint,which has obtained satisfactory clinical efficacy,but the postoperative periprosthetic osteolysis (PPO),which is mediated by wear particles,restricts the long-term effect of artificial joints.It is found that wear particles increase the expression of cytokines and inflammatory factors by stimulating the cells around the prosthesis,activate different signaling pathways,promote the imbalance between bone formation mediated by local osteoblasts and bone resorption mediated by osteoclast so as to lose of the local bone mass,and eventually produce osteolysis and aseptic loosening.This article reviews the different signal pathways activated by wear particles in recent years,in order to explore the pathogenesis of PPO and to open up new avenues for its prevention and treatment.