Dry eye disease is a chronic ocular surface disorder of multifactorial origin, characterized by a loss of tear film homeostasis and associated with a range of ocular discomfort symptoms. Growing evidence underscores a significant bidirectional relationship between dry eye and depression: individuals with dry eye disease exhibit a higher prevalence of depressive disorders, and conversely, those diagnosed with depression demonstrate an increased susceptibility to developing dry eye. This interplay is mediated through several pathophysiological pathways, such as chronic inflammation, cerebral functional alterations, gut microbiome dysregulation, and sleep disturbances, which may collectively sustain a vicious cycle. The use of antidepressant therapy introduces further complexity, exerting heterogeneous effects on dry eye—some agents may offer symptomatic relief, whereas others can aggravate ocular surface impairment. The mechanisms responsible for these differential outcomes remain incompletely elucidated and merit further investigation. This review systematically consolidates epidemiological data on the dry eye-depression link, examines potential shared pathological mechanisms, and evaluates current therapeutic options. We propose an integrated management approach that combines conventional dry eye treatments, such as traditional Chinese medicine, electroacupuncture, physical activity and antidepressants—a multimodal strategy that may yield synergistic benefits in alleviating both ocular and affective symptoms, thereby improving overall quality of life. Moving forward, research should focus on deciphering the underlying mechanistic pathways and facilitating the translation of these insights into clinical practice to inform targeted, combined treatment regimens for patients with dry eye and depression.

Traditional development of small protein scaffolds has relied on display technologies and mutation-based engineering, which limit sequence and functional diversity, thereby constraining their therapeutic and application potential. Protein design tools have significantly advanced the creation of novel protein sequences, structures, and functions. However, further improvements in design strategies are still needed to more efficiently optimize the functional performance of protein-based drugs and enhance their druggability. Here, we extended an evolution-based design protocol to create a novel minibinder, BindHer, against the human epidermal growth factor receptor 2 (HER2). It not only exhibits super stability and binding selectivity but also demonstrates remarkable properties in tissue specificity. Radiolabeling experiments with ^99mTc, ⁶⁸Ga, and ¹⁸F revealed that BindHer efficiently targets tumors in HER2-positive breast cancer mouse models, with minimal nonspecific liver absorption, outperforming scaffolds designed through traditional engineering. These findings highlight a new rational approach to automated protein design, offering significant potential for large-scale applications in therapeutic mini-protein development.

OBJECTIVES: To investigate the risk factors of overall postoperative complications in elderly patients undergoing gastrointestinal surgeries. METHODS: This study was conducted among a total of 1388 elderly patients, who underwent elective gastrointestinal surgeries at 17 centers across China between April, 2020 and April, 2022. The primary outcome was the incidence of postoperative complications within 30 days, including procedure-related, neuropsychiatric, respiratory, cardiovascular, and gastrointestinal complications as well as acute kidney injury. Baseline characteristics, preoperative psychological and functional status, intraoperative anesthesia and surgical factors, intraoperative medication, use of nerve block, and postoperative analgesia methods were compared between the patients experiencing one or more postoperative complications and those without complications. Univariate and multivariate logistic regression analyses were performed to identify the independent risk factors for postoperative complications. The relationship between postoperative acute pain and each type of complication were explored. RESULTS: The incidence of overall postoperative complications was 50.8% (705/1388) in these patients. Multivariate analysis showed that age (OR: 1.026; 95% CI: 1.006-1.046), prognostic nutritional index (OR: 0.998; 95% CI: 0.997-1.000), preoperative EuroQol-5 dimensions score (OR: 0.094; 95% CI: 0.018-0.500), blood loss (OR: 1.002; 95% CI: 1.001-1.003), and acute postoperative pain (OR: 1.308; 95% CI: 1.033-1.657) were significantly associated with the occurrence of postoperative complications. Specifically, patients experiencing severe postoperative pain had a significantly higher incidence of neuropsychiatric (27.2% vs 19.8%), procedure-related (17.3% vs 10.2%), and cardiovascular complications (3.6% vs 1.7%). CONCLUSIONS An advanced age, a low preoperative nutritional index, a poor quality of life score, a greater volume of intraoperative blood loss, and acute postoperative pain are independent risk factors for postoperative complications in elderly patients undergoing gastrointestinal surgeries. There is a significant association between acute postoperative pain and multi-system complications.
Humans ; Postoperative Complications/etiology* ; Aged ; Risk Factors ; Digestive System Surgical Procedures/adverse effects* ; Male ; Female ; China/epidemiology* ; Pain, Postoperative/epidemiology* ; Incidence ; Aged, 80 and over

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 analyze the electrocardiographic characteristics of patients with chronic pulmonary artery stenosis (PAS), and to explore their relationship with disease severity indicators.Methods:The study was a retrospective case-series analysis. Patients with chronic PAS admitted to Gansu Provincial Hospital from January 2018 to July 2021 were enrolled. The clinical data and the results of electrocardiography, transthoracic echocardiography, right cardiac catheterization, N-terminal B-type natriuretic peptide (NT-proBNP) measurement and 6-min walking distance test of patients were analyzed. The linear regression model or logistic regression model was used to analyze the relationship between electrocardiographic characteristics and the disease severity in patients with chronic PAS.Results:Sixty-three patients aged (62.1±9.7) years including 43 females (68.3%) were enrolled in the study. Among them, 62 patients (98.4%) had (R 1+S Ⅲ)-(S Ⅰ+R Ⅲ)<1.5 mV, and no patients had V 5lead R: S ratio to V 1 lead R: S<0.04 and V 6 lead R: S ratio<0.4. There were 55 patients (87.3%), with flat or inverted T-waves in V 1, and 10 patients (15.9%) with flat or inverted T-waves in all precordial leads (V 1-V 6). There were 18 patients (28.6%) with flat or inverted T-waves in inferior leads (Ⅱ, Ⅲ, aVF). Multiple liner regression analysis showed that Max R V1, 2+Max S I, aVL-S V1 combined with the number of flat or inverted T-waves in limb leads was independently correlated with atrial area ( R2=0.290, P=0.002); R V1+S V5 was independently correlated with right ventricular area ( R2=0.257, P=0.001); R peak V 1 combined with the number of flat or inverted T waves in precordial leads was independently correlated with tricuspid annular plane systolic excursion ( R2=0.407, P<0.001); (R 1+S Ⅲ)-(S Ⅰ+R Ⅲ) combined with the number of flat or inverted T waves in precordial leads was independently correlated with NT-proBNP ( R2=0.504, P<0.001); Max R V1, 2+Max S I, aVL-S V1 were independently correlated with right atrial pressure ( R2=0.803, P=0.036); (R 1+S Ⅲ)-(S Ⅰ+R Ⅲ) were independently correlated with mean pulmonary artery pressure ( R2=0.302, P<0.001); R aVRcombined with the number of flat or inverted T-waves in precordial leads was independently correlated with cardiac index ( R2=0.173, P=0.003); (R 1+S Ⅲ)-(S Ⅰ+R Ⅲ) was independently correlated with pulmonary vascular resistance ( R2=0.173, P=0.002); R peak V 1 combined with the number of flat or inverted T-waves in precordial leads was independently correlated with mixed vein oxygen saturation ( R2=0.302, P<0.001). Conclusion:The vast majority of patients with chronic PAS have (R 1+S Ⅲ)-(S Ⅰ+R Ⅲ)<1.5 mV and flat or inverted T-wave in V 1 lead, and some characteristic electrocardiographic manifestations are correlated with indicators of disease severity.