The specific mechanisms of interferon regulatory factor 8(IRF8)in porcine intestinal in-nate immunity and resistance to enteric virus infection remain to be elucidated.To investigate the immunoregulatory role of IRF8,establishing an IRF8 gene knockout porcine intestinal epithelial cell(IPEC-J2)monoclonal cell line is of significant importance.This study initially aimed to obtain recombinant adeno-associated virus rAAV-sgIRF8-eGFP capable of knocking out the IRF8 gene through co-transfection of HEK-293T cells with three plasmids.Subsequently,IPEC-J2 cells were infected with the virus,and those expressing eGFP were selected by flow cytometry and cultured to form monoclonal cell lines.These cell lines were then identified by Sanger sequencing and West-ern blot techniques.Lastly,qPCR analysis was used to measure the expression levels of interferon factors IFN-α,IFN-β,IFN-γ and IFN-λ,providing preliminary insights into the impact of IRF8 gene knockout on IPEC-J2 cell immunity.The results demonstrated successful generation of rAAV-sgIRF8-eGFP,which successfully infected IPEC-J2 cells leading to eGFP fluorescence.Flow cytometry followed by cell culture led to the establishment of two monoclonal cell lines,IRF8-KO1 and IRF8-KO3.Sanger sequencing revealed a five-base deletion in IRF8-KO1 and a seven-base dele-tion in IRF8-KO3.Western blot confirmed the absence of IRF8 protein expression in IRF8-KO1,making it an ideal candidate monoclonal cell line.qPCR analysis of interferon factors indicated sig-nificant decrease in IFN-γ(P＜0.05)and IFN-λ(P＜0.01)transcription level in IRF8-knockout cells,while the transcription levels of IFN-α and IFN-β remained relatively unchanged.This study successfully established an IRF8 gene knockout IPEC-J2 monoclonal cell line,providing a founda-tion for further research on IRF8-related porcine intestinal immune regulation and mechanisms of intestinal virus infection.

Objective To investigate the role of circular RNAs(circRNA)in Alzheimer's disease(AD)and its potential mechanism.Methods Six-month-old APP/PS1 mouse model of AD and wild type(WT)mice were subjected and then randomly divided into WT group,WT+circ-Olfm1 knockout group,AD group(transgenic APP/PS1 mice),AD+circ-Olfm1 knockout group,AD+FOXO3a knockout group,with 3 mice in each group.① The total RNA of mouse brain was extracted,and the differential expression of circRNAs and mRNAs between the AD mice and WT mice was detected,and the obtained circRNAs and mRNAs were analyzed with gene ontology(GO)analysis.② RT-qPCR was used to detect the expression of the top 10 up-regulated and down-regulated circRNAs,as well as the expression of circ-Olfm1 and miR-330-5p.③ Lentiviral vectors were prepared and stereotaxically injected into the cortex or hippocampus of WT and AD mice to knock out circ-Olfm1 gene.Water maze test was used to evaluate the effect of circ-Olfm1 knockout on cognitive function,and immunofluorescence assay was employed to observe the deposition of amyloid β(Aβ)plaque in the brain.④ The interaction between circ-Olfm1 and miR-330-5p was verified by double luciferase reporter gene analysis.⑤ The protein levels of AMPK and FOXO3a were detected by Western blotting.⑥ Transmission electron microscopy was utilized to observe the mitochondria of the hippocampus.⑦ The levels of inflammatory factors IL-6,IL-1β and TNF-α were detected by ELISA.Results There were totally 52 differentially expressed circRNAs identified between the AD and WT mice,including 28 up-regulated and 24 down-regulated(fold change>1.5,P<0.05).These differentially expressed genes are mainly involved in signal transduction,learning and memory and other functions.circ-Olfm1 was identified as the most significantly differentially expressed circRNA,which is highly expressed in the neurons and up-regulated in the cerebral cortex and hippocampus of the AD mice.Knockout of circ-Olfm1 reduced the number of Aβ plaques in the cerebral cortex and hippocampus of AD mice(P<0.01).In starBase database,there are complementary sequences observed between circ-Olfm1 and miR-330-5p.Western blotting showed that the addition of Aβ42 significantly increased the expression of AMPK and FOXO3a in the neuronal cells(P<0.01).And silencing circ-Olfm1 led to decreased expression of AMPK and FOXO3a in neuronal cells+Aβ42(P<0.01).ELISA revealed that knockout of FOXO3a significantly increased the levels of inflammatory factors IL-6,IL-1β,and TNF-α(P<0.01).Transmission electron microscopy displayed that knocking FOXO3a out significantly aggravated mitochondrial damage(P<0.01).Conclusion circ-Olfm1 is up-regulated in the brain tissue and neurons+Aβ42 of AD rats,and the mechanism of cognitive impairment in AD rats may be through its regulating FOXO3a protein.

The specific mechanisms of interferon regulatory factor 8(IRF8)in porcine intestinal in-nate immunity and resistance to enteric virus infection remain to be elucidated.To investigate the immunoregulatory role of IRF8,establishing an IRF8 gene knockout porcine intestinal epithelial cell(IPEC-J2)monoclonal cell line is of significant importance.This study initially aimed to obtain recombinant adeno-associated virus rAAV-sgIRF8-eGFP capable of knocking out the IRF8 gene through co-transfection of HEK-293T cells with three plasmids.Subsequently,IPEC-J2 cells were infected with the virus,and those expressing eGFP were selected by flow cytometry and cultured to form monoclonal cell lines.These cell lines were then identified by Sanger sequencing and West-ern blot techniques.Lastly,qPCR analysis was used to measure the expression levels of interferon factors IFN-α,IFN-β,IFN-γ and IFN-λ,providing preliminary insights into the impact of IRF8 gene knockout on IPEC-J2 cell immunity.The results demonstrated successful generation of rAAV-sgIRF8-eGFP,which successfully infected IPEC-J2 cells leading to eGFP fluorescence.Flow cytometry followed by cell culture led to the establishment of two monoclonal cell lines,IRF8-KO1 and IRF8-KO3.Sanger sequencing revealed a five-base deletion in IRF8-KO1 and a seven-base dele-tion in IRF8-KO3.Western blot confirmed the absence of IRF8 protein expression in IRF8-KO1,making it an ideal candidate monoclonal cell line.qPCR analysis of interferon factors indicated sig-nificant decrease in IFN-γ(P＜0.05)and IFN-λ(P＜0.01)transcription level in IRF8-knockout cells,while the transcription levels of IFN-α and IFN-β remained relatively unchanged.This study successfully established an IRF8 gene knockout IPEC-J2 monoclonal cell line,providing a founda-tion for further research on IRF8-related porcine intestinal immune regulation and mechanisms of intestinal virus infection.

Objective:To explore the effectiveness of a hysteroscopic endometrial lesion diagnosis model de-veloped based on deep learning(DL)algorithm combined with gradient-weighted class activation mapping(Grad-CAM)visualization technology.Methods:303 hysteroscopy videos(4781 images)of 291 patients who un-derwent hysteroscopy examination in the Department of Gynecology,Renmin Hospital of Wuhan University from June 1,2021 to December 31,2022 were selected.The dataset was divided into a training set(3703 images)and a test set(1078 images)by weight sampling method.After the training set was used for model learning and train-ing,two model architectures,residual neural network(ResNet18)and efficient neural network(EfficientNet-B0),were selected to verify the model in the test set by five-class and two-class classification tasks,respectively.Tak-ing histopathology as the gold standard,the diagnostic efficacy was evaluated to select the optimal model,and the Grad-CAM layer was embedded in the optimal model to output hysteroscopy images of Grad-CAM.Results:①In the five-class classification tasks,the accuracy of EfficientNet-B0 model(93.23％)was higher than that of Res-Net18 model(84.23％);the area under the curve(AUC)of EfficientNet-B0 model in the diagnosis of five disea-ses,including atypical endometrial hyperplasia,endometrial polyps,endometrial cancer,endometrial atypical hy-perplasia,and submucous myoma,was slightly higher than that of ResNet18 model,and the AUC of both models was almost above 0.980.②In the binary classification task of accuracy and the evaluation of specificity,the two models were similar,both above 93.00％,and the sensitivity of EfficientNet-B0 model(91.14％)was significantly better than that of ResNet18 model(77.22％).③EfficientNet-B0 model combined with Grad-CAM algorithm could identify the abnormal areas in the image.After biopsy and pathological examination,it was confirmed that about 95％of the marked areas in the model's output heatmap were lesion areas.Conclusions:The hysteroscopy di-agnostic model developed by EfficientNet-B0 model combined with Grad-CAM has high diagnostic accuracy,sen-sitivity,and specificity,and has application value in the diagnosis of endometrial lesions.

Objective:To explore the effectiveness of a hysteroscopic endometrial lesion diagnosis model de-veloped based on deep learning(DL)algorithm combined with gradient-weighted class activation mapping(Grad-CAM)visualization technology.Methods:303 hysteroscopy videos(4781 images)of 291 patients who un-derwent hysteroscopy examination in the Department of Gynecology,Renmin Hospital of Wuhan University from June 1,2021 to December 31,2022 were selected.The dataset was divided into a training set(3703 images)and a test set(1078 images)by weight sampling method.After the training set was used for model learning and train-ing,two model architectures,residual neural network(ResNet18)and efficient neural network(EfficientNet-B0),were selected to verify the model in the test set by five-class and two-class classification tasks,respectively.Tak-ing histopathology as the gold standard,the diagnostic efficacy was evaluated to select the optimal model,and the Grad-CAM layer was embedded in the optimal model to output hysteroscopy images of Grad-CAM.Results:①In the five-class classification tasks,the accuracy of EfficientNet-B0 model(93.23％)was higher than that of Res-Net18 model(84.23％);the area under the curve(AUC)of EfficientNet-B0 model in the diagnosis of five disea-ses,including atypical endometrial hyperplasia,endometrial polyps,endometrial cancer,endometrial atypical hy-perplasia,and submucous myoma,was slightly higher than that of ResNet18 model,and the AUC of both models was almost above 0.980.②In the binary classification task of accuracy and the evaluation of specificity,the two models were similar,both above 93.00％,and the sensitivity of EfficientNet-B0 model(91.14％)was significantly better than that of ResNet18 model(77.22％).③EfficientNet-B0 model combined with Grad-CAM algorithm could identify the abnormal areas in the image.After biopsy and pathological examination,it was confirmed that about 95％of the marked areas in the model's output heatmap were lesion areas.Conclusions:The hysteroscopy di-agnostic model developed by EfficientNet-B0 model combined with Grad-CAM has high diagnostic accuracy,sen-sitivity,and specificity,and has application value in the diagnosis of endometrial lesions.

Objective:To explore the effectiveness of a hysteroscopic endometrial lesion diagnosis model de-veloped based on deep learning(DL)algorithm combined with gradient-weighted class activation mapping(Grad-CAM)visualization technology.Methods:303 hysteroscopy videos(4781 images)of 291 patients who un-derwent hysteroscopy examination in the Department of Gynecology,Renmin Hospital of Wuhan University from June 1,2021 to December 31,2022 were selected.The dataset was divided into a training set(3703 images)and a test set(1078 images)by weight sampling method.After the training set was used for model learning and train-ing,two model architectures,residual neural network(ResNet18)and efficient neural network(EfficientNet-B0),were selected to verify the model in the test set by five-class and two-class classification tasks,respectively.Tak-ing histopathology as the gold standard,the diagnostic efficacy was evaluated to select the optimal model,and the Grad-CAM layer was embedded in the optimal model to output hysteroscopy images of Grad-CAM.Results:①In the five-class classification tasks,the accuracy of EfficientNet-B0 model(93.23％)was higher than that of Res-Net18 model(84.23％);the area under the curve(AUC)of EfficientNet-B0 model in the diagnosis of five disea-ses,including atypical endometrial hyperplasia,endometrial polyps,endometrial cancer,endometrial atypical hy-perplasia,and submucous myoma,was slightly higher than that of ResNet18 model,and the AUC of both models was almost above 0.980.②In the binary classification task of accuracy and the evaluation of specificity,the two models were similar,both above 93.00％,and the sensitivity of EfficientNet-B0 model(91.14％)was significantly better than that of ResNet18 model(77.22％).③EfficientNet-B0 model combined with Grad-CAM algorithm could identify the abnormal areas in the image.After biopsy and pathological examination,it was confirmed that about 95％of the marked areas in the model's output heatmap were lesion areas.Conclusions:The hysteroscopy di-agnostic model developed by EfficientNet-B0 model combined with Grad-CAM has high diagnostic accuracy,sen-sitivity,and specificity,and has application value in the diagnosis of endometrial lesions.

Objective:To explore the effectiveness of a hysteroscopic endometrial lesion diagnosis model de-veloped based on deep learning(DL)algorithm combined with gradient-weighted class activation mapping(Grad-CAM)visualization technology.Methods:303 hysteroscopy videos(4781 images)of 291 patients who un-derwent hysteroscopy examination in the Department of Gynecology,Renmin Hospital of Wuhan University from June 1,2021 to December 31,2022 were selected.The dataset was divided into a training set(3703 images)and a test set(1078 images)by weight sampling method.After the training set was used for model learning and train-ing,two model architectures,residual neural network(ResNet18)and efficient neural network(EfficientNet-B0),were selected to verify the model in the test set by five-class and two-class classification tasks,respectively.Tak-ing histopathology as the gold standard,the diagnostic efficacy was evaluated to select the optimal model,and the Grad-CAM layer was embedded in the optimal model to output hysteroscopy images of Grad-CAM.Results:①In the five-class classification tasks,the accuracy of EfficientNet-B0 model(93.23％)was higher than that of Res-Net18 model(84.23％);the area under the curve(AUC)of EfficientNet-B0 model in the diagnosis of five disea-ses,including atypical endometrial hyperplasia,endometrial polyps,endometrial cancer,endometrial atypical hy-perplasia,and submucous myoma,was slightly higher than that of ResNet18 model,and the AUC of both models was almost above 0.980.②In the binary classification task of accuracy and the evaluation of specificity,the two models were similar,both above 93.00％,and the sensitivity of EfficientNet-B0 model(91.14％)was significantly better than that of ResNet18 model(77.22％).③EfficientNet-B0 model combined with Grad-CAM algorithm could identify the abnormal areas in the image.After biopsy and pathological examination,it was confirmed that about 95％of the marked areas in the model's output heatmap were lesion areas.Conclusions:The hysteroscopy di-agnostic model developed by EfficientNet-B0 model combined with Grad-CAM has high diagnostic accuracy,sen-sitivity,and specificity,and has application value in the diagnosis of endometrial lesions.

Objective:To explore the effectiveness of a hysteroscopic endometrial lesion diagnosis model de-veloped based on deep learning(DL)algorithm combined with gradient-weighted class activation mapping(Grad-CAM)visualization technology.Methods:303 hysteroscopy videos(4781 images)of 291 patients who un-derwent hysteroscopy examination in the Department of Gynecology,Renmin Hospital of Wuhan University from June 1,2021 to December 31,2022 were selected.The dataset was divided into a training set(3703 images)and a test set(1078 images)by weight sampling method.After the training set was used for model learning and train-ing,two model architectures,residual neural network(ResNet18)and efficient neural network(EfficientNet-B0),were selected to verify the model in the test set by five-class and two-class classification tasks,respectively.Tak-ing histopathology as the gold standard,the diagnostic efficacy was evaluated to select the optimal model,and the Grad-CAM layer was embedded in the optimal model to output hysteroscopy images of Grad-CAM.Results:①In the five-class classification tasks,the accuracy of EfficientNet-B0 model(93.23％)was higher than that of Res-Net18 model(84.23％);the area under the curve(AUC)of EfficientNet-B0 model in the diagnosis of five disea-ses,including atypical endometrial hyperplasia,endometrial polyps,endometrial cancer,endometrial atypical hy-perplasia,and submucous myoma,was slightly higher than that of ResNet18 model,and the AUC of both models was almost above 0.980.②In the binary classification task of accuracy and the evaluation of specificity,the two models were similar,both above 93.00％,and the sensitivity of EfficientNet-B0 model(91.14％)was significantly better than that of ResNet18 model(77.22％).③EfficientNet-B0 model combined with Grad-CAM algorithm could identify the abnormal areas in the image.After biopsy and pathological examination,it was confirmed that about 95％of the marked areas in the model's output heatmap were lesion areas.Conclusions:The hysteroscopy di-agnostic model developed by EfficientNet-B0 model combined with Grad-CAM has high diagnostic accuracy,sen-sitivity,and specificity,and has application value in the diagnosis of endometrial lesions.

Objective:To explore the effectiveness of a hysteroscopic endometrial lesion diagnosis model de-veloped based on deep learning(DL)algorithm combined with gradient-weighted class activation mapping(Grad-CAM)visualization technology.Methods:303 hysteroscopy videos(4781 images)of 291 patients who un-derwent hysteroscopy examination in the Department of Gynecology,Renmin Hospital of Wuhan University from June 1,2021 to December 31,2022 were selected.The dataset was divided into a training set(3703 images)and a test set(1078 images)by weight sampling method.After the training set was used for model learning and train-ing,two model architectures,residual neural network(ResNet18)and efficient neural network(EfficientNet-B0),were selected to verify the model in the test set by five-class and two-class classification tasks,respectively.Tak-ing histopathology as the gold standard,the diagnostic efficacy was evaluated to select the optimal model,and the Grad-CAM layer was embedded in the optimal model to output hysteroscopy images of Grad-CAM.Results:①In the five-class classification tasks,the accuracy of EfficientNet-B0 model(93.23％)was higher than that of Res-Net18 model(84.23％);the area under the curve(AUC)of EfficientNet-B0 model in the diagnosis of five disea-ses,including atypical endometrial hyperplasia,endometrial polyps,endometrial cancer,endometrial atypical hy-perplasia,and submucous myoma,was slightly higher than that of ResNet18 model,and the AUC of both models was almost above 0.980.②In the binary classification task of accuracy and the evaluation of specificity,the two models were similar,both above 93.00％,and the sensitivity of EfficientNet-B0 model(91.14％)was significantly better than that of ResNet18 model(77.22％).③EfficientNet-B0 model combined with Grad-CAM algorithm could identify the abnormal areas in the image.After biopsy and pathological examination,it was confirmed that about 95％of the marked areas in the model's output heatmap were lesion areas.Conclusions:The hysteroscopy di-agnostic model developed by EfficientNet-B0 model combined with Grad-CAM has high diagnostic accuracy,sen-sitivity,and specificity,and has application value in the diagnosis of endometrial lesions.

Objective:To explore the effectiveness of a hysteroscopic endometrial lesion diagnosis model de-veloped based on deep learning(DL)algorithm combined with gradient-weighted class activation mapping(Grad-CAM)visualization technology.Methods:303 hysteroscopy videos(4781 images)of 291 patients who un-derwent hysteroscopy examination in the Department of Gynecology,Renmin Hospital of Wuhan University from June 1,2021 to December 31,2022 were selected.The dataset was divided into a training set(3703 images)and a test set(1078 images)by weight sampling method.After the training set was used for model learning and train-ing,two model architectures,residual neural network(ResNet18)and efficient neural network(EfficientNet-B0),were selected to verify the model in the test set by five-class and two-class classification tasks,respectively.Tak-ing histopathology as the gold standard,the diagnostic efficacy was evaluated to select the optimal model,and the Grad-CAM layer was embedded in the optimal model to output hysteroscopy images of Grad-CAM.Results:①In the five-class classification tasks,the accuracy of EfficientNet-B0 model(93.23％)was higher than that of Res-Net18 model(84.23％);the area under the curve(AUC)of EfficientNet-B0 model in the diagnosis of five disea-ses,including atypical endometrial hyperplasia,endometrial polyps,endometrial cancer,endometrial atypical hy-perplasia,and submucous myoma,was slightly higher than that of ResNet18 model,and the AUC of both models was almost above 0.980.②In the binary classification task of accuracy and the evaluation of specificity,the two models were similar,both above 93.00％,and the sensitivity of EfficientNet-B0 model(91.14％)was significantly better than that of ResNet18 model(77.22％).③EfficientNet-B0 model combined with Grad-CAM algorithm could identify the abnormal areas in the image.After biopsy and pathological examination,it was confirmed that about 95％of the marked areas in the model's output heatmap were lesion areas.Conclusions:The hysteroscopy di-agnostic model developed by EfficientNet-B0 model combined with Grad-CAM has high diagnostic accuracy,sen-sitivity,and specificity,and has application value in the diagnosis of endometrial lesions.