Objective To study the application value of 3.0T MRI T2 mapping quantitative technology in the diagnosis of anterior cruciate ligament sprain and chronic injury of knee joint.Methods A total of 82 subjects were studied,and the experimental group 72 cases was divided into grade Ⅰ injury group(25 cases),grade Ⅱ injury group(25 cases),chronic injury group(22 cases),and control group 10 cases.The experimental group met the criteria of arthroscopy.The proximal,middle,and distal segments of the anterior cruciate ligament were selected as the region of interest(ROI),and T2 mapping values were measured.The differences in T2 mapping values of each area were compared between and within the groups,while compared with arthroscopy.Results The T2 mapping values in grade Ⅰ,Ⅱ,and chronic injury groups were higher than those in control group(P<0.05).Comparison within the experimental group:the T2 mapping values of each area in grade Ⅱ injury group were higher than those in grade Ⅰ injury group and chronic injury group(P<0.05).The T2 mapping values of each area in grade Ⅰ injury group were higher than those in chronic injury group(P<0.05).The specificity,sensitivity,positive predictive value,negative predictive value and accuracy of T2 mapping in diagnosing anterior cruciate ligament grade Ⅰ injury were 94.7%,95.5%,89.7%,96.6%,and 90.2%respectively.The specificity,sensitivity,positive predictive value,negative predictive value,and accuracy of grade Ⅱ injury were 89.4%,87.9%,92.1%,93.4%,and 93.8%respectively.The specificity,sensitivity,positive predictive value,negative predictive value,and accuracy of chronic injury were 92.2%,95.4%,90.3%,87.6%,and 91.5%respectively.Kappa test showed a good con-sistency between T2 mapping results and arthroscopic results,with a Kappa value of 0.763(P<0.01).Conclusion The value of MRI T2 mapping can provide a reference for the clinical diagnosis of anterior cruciate ligament sprain and chronic injury of knee joint,and the results are in good agreement with the control of arthroscopy.

ObjectiveTo explore the therapeutic effect and mechanism of Guipitang on rats with myocardial ischemia. MethodFifty SD rats were divided into five groups： a control group， a model group， low and high-dose Guipitang （7.52， 15.04 g·kg^-1） groups， and a trimetazidine group （0.002 g·kg^-1）. By intragastric administration of vitamin D₃ and feeding rats with high-fat forage and injecting isoproterenol， the rat model of myocardial ischemia was established. After drug treatment of 15 d， an electrocardiogram （ECG） was performed to analyze the degree of myocardial injury. A fully automatic biochemical analyzer was used to detect the changes in the serum levels of total cholesterol （TC）， high-density lipoprotein cholesterol （HDL-C）， triglyceride （TG）， and low-density lipoprotein cholesterol （LDL-C）. Hematoxylin-eosin （HE） staining and Masson staining were used to observe myocardial histopathological changes. TdT-mediated dUTP nick end labeling （TUNEL） staining was used to detect cardiomyocyte apoptosis. Western blot was adopted to detect the protein levels of extracellular signal-regulated kinase 1/2 （ERK1/2）， phospho-ERK1/2 （p-ERK1/2）， p38 mitogen-activated protein kinase （p38 MAPK）， phospho-p38 MAPK （p-p38 MAPK）， B-cell lymphoma-2 （Bcl-2）-associated X （Bax）， Bcl-2， and cleaved cysteine aspartate proteolytic enzyme （cleaved Caspase-3）. ResultCompared with the control group， the ECG S-T segment decreased in the model group. The serum levels of TC， TG， and LDL-C were increased significantly （P<0.05）. The arrangement of myocardial tissue was disordered， and the proportion of cardiomyocyte apoptosis increased. The protein levels of cleaved Caspase-3， Bax， and p-p38 MAPK in the heart were increased， and the Bcl-2 expression was decreased （P<0.05）. Compared with the model group， the S-T segment downward shift was restored in the low and high-dose Guipitang groups and trimetazidine group， and the levels of TC， TG， and LDL-C were decreased. The protein expression of cleaved Caspase-3 and Bax in the heart dropped， and p-p38 MAPK and p-ERK1/2 protein expressions increased significantly （P<0.05）. The degree of myocardial injury was alleviated， and the proportion of cardiomyocyte apoptosis decreased. Bcl-2 protein expression was increased significantly in the low-dose Guipitang group （P<0.05）. ERK1/2 and p38 MAPK proteins had no significant difference among different groups. ConclusionGuipitang could alleviate myocardial injury and inhibit cardiomyocyte apoptosis in rats by activating the expression of ERK1/2 and p38 MAPK.

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 compare the efficacy of mediastinoscope-assisted transhiatal esophagectomy (MATHE) and functional minimally invasive esophagectomy (FMIE) for esophageal cancer. Methods Patients who underwent minimally invasive esophagectomy at Jining No.1 Hospital from March 2018 to September 2022 were retrospectively included. The patients were divided into a MATHE group and a FMIE group according to the procedures. The patients were matched via propensity score matching (PSM) with a ratio of 1 : 1 and a caliper value of 0.2. The clinical data of the patients were compared after the matching. Results A total of 73 patients were include in the study, including 54 males and 19 females, with an average age of (65.12±7.87) years. There were 37 patients in the MATHE group and 36 patients in the FMIE group. Thirty pairs were successfully matched. Compared with the FMIE group, MATHE group had shorter operation time (P=0.022), lower postoperative 24 h pain score (P=0.031), and less drainage on postoperative 1-3 days (P<0.001). FMIE group had more lymph node dissection (P<0.001), lower incidence of postoperative hoarseness (P=0.038), lower white blood cell and neutrophil counts on postoperative 1 day (P<0.001). There was no statistically significant difference in the bleeding volume, R0 resection, hospital mortality, postoperative hospital stay, anastomotic leak, chylothorax, or pulmonary infection between the two groups (P>0.05). Conclusion Compared with the FMIE, MATHE has shorter operation time, less postoperative pain and drainage, but removes less lymph nodes, which is deficient in oncology. For some special patients such as those with early cancer or extensive pleural adhesions, MATHE may be a suitable surgical method.

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.