ObjectiveTo identify the patients with metabolic dysfunction-associated fatty liver disease （MAFLD） among the health check-up population， and to perform stratified management of patients with the low， medium， and high risk of advanced fibrosis based on noninvasive fibrosis scores. MethodsA cross-sectional study was conducted among 3 125 individuals who underwent physical examination in Beijing Physical Examination Center from December 2017 to December 2019， and they were divided into MAFLD group with 1 068 individuals and non-MAFLD group with 2 057 individuals. According to BMI， the MAFLD group was further divided into lean MAFLD group （125 individuals with BMI<24 kg/m²） and non-lean MAFLD group （943 individuals with BMI≥24 kg/m²）. Indicators including demographic data， past history， laboratory examination， and liver ultrasound were compared between groups. Fibrosis-4 （FIB-4） score， NAFLD fibrosis score （NFS）， aspartate aminotransferase-to-platelet ratio index （APRI）， and BARD score were calculated for the patients in the MAFLD group to assess the risk of advanced fibrosis. The independent-samples t test was used for comparison of normally distributed continuous data between two groups， and the Mann-Whitney U rank sum test was used for comparison of non-normally distributed continuous data between two groups； the chi-square test or the Fisher’s exact test was used for comparison of categorical data between groups. A logistic regression analysis was used to investigate the influence of each indicator in MAFLD. ResultsCompared with the non-MAFLD group， the MAFLD group had significantly higher age （Z=-9.758， P<0.05）， proportion of male patients （χ²=137.555， P<0.05）， and levels of body weight （Z=-27.987， P<0.05）， BMI （Z=-32.714， P<0.05）， waist circumference （Z=-31.805， P<0.05）， hip circumference （Z=-26.342， P<0.05）， waist-hip ratio （Z=-28.554， P<0.05）， alanine aminotransferase （ALT） （Z=-25.820， P<0.05）， aspartate aminotransferase （AST） （Z=-16.894， P<0.05）， gamma-glutamyl transpeptidase （GGT） （Z=-25.069， P<0.05）， alkaline phosphatase （Z=-12.533， P<0.05）， triglyceride （Z=-27.559）， total cholesterol （Z=-7.833， P<0.05）， low-density lipoprotein cholesterol （LDL-C） （Z=-8.222， P<0.05）， and uric acid （UA） （Z=-20.024， P<0.05）， as well as a significantly higher proportion of patients with metabolic syndrome （MetS） （χ²=578.220， P<0.05）， significantly higher prevalence rates of hypertension （χ²=241.694， P<0.05）， type 2 diabetes （χ²=796.484， P<0.05）， and dyslipidemia （χ²=369.843， P<0.05）， and a significant reduction in high-density lipoprotein cholesterol （HDL-C） （Z=23.153， P<0.001）. The multivariate logistic regression analysis showed that male sex （odds ratio ［OR］=1.45， 95% confidence interval ［CI］： 1.203‍ ‍—‍ ‍1.737）， ALT （OR=1.05， 95%CI： 1.046‍ ‍—‍ ‍1.062）， LDL-C （OR=1.23， 95%CI： 1.102‍ ‍—‍ ‍1.373）， and comorbidity with MetS （OR=5.97， 95%CI： 4.876‍ ‍—‍ ‍7.316） were independently associated with MAFLD. Compared with the non-lean MAFLD group， the lean MAFLD group had significantly higher age （Z=3.736， P<0.05） and HDL-C （Z=2.679， P<0.05） and significant reductions in the proportion of male patients （χ²=28.970， P<0.05）， body weight （Z=-14.230， P<0.05）， BMI （Z=-18.188， P<0.05）， waist circumference （Z=-13.451， P<0.05）， hip circumference （Z=-13.317， P<0.05）， ALT （Z=-4.519， P<0.05）， AST （Z=-2.258， P<0.05）， GGT （Z=-4.592， P<0.05）， UA （Z=-4.415， P<0.05）， the proportion of patients with moderate or severe fatty liver disease or MetS （χ²=42.564， P<0.05）， and the prevalence rates of hypertension （χ²=12.057， P<0.05） and type 2 diabetes （χ²=3.174， P<0.05）. Among the patients with MAFLD， 10 patients （0.9%） had an FIB-4 score of >2.67， 4 patients （0.4%） had an NFS score of >0.676， 8 patients （0.7%） had an APRI of >1， and 551 patients （51.6%） had a BARD score of ≥2. ConclusionThere is a relatively high prevalence rate of MAFLD among the health check-up population in Beijing， but with a relatively low number of patients with a high risk of advanced fibrosis， and such patients need to be referred to specialized hospitals for liver diseases.

Objective: To analyze the current status and influencing factors of psychological crisis among college students, so as to provide a scientific basis for the formulation of psychological crisis intervention plans in colleges and universities. Methods: From September to December 2024, 645 college students from a medical undergraduate university in Heilongjiang Province were selected with a convenience sampling method. A convergent mixed analysis design was used. Quantitative analysis was conducted using College Students Psychological Crisis Screening Scale, Emotion Regulation Questionnaire, Short-Egna Minnen av Barndoms Uppfostran and Perceived School Climate Scale. Binary Logistic regression analysis was used to explore the related factors of psychological crisis among college students. Qualitative research was conducted on 15 college students with psychological crisis identified in the quantitative analysis by a purposive sampling method. The interview data were organized and analyzed using the thematic framework analysis method. Results: Among the surveyed college students, 92 (14.3%) had psychological crisis. Binary Logistic regression analysis results showed that positive parenting style ( OR=0.97,95%CI =0.95-0.99), negative parenting style ( OR=1.01,95%CI =1.00-1.02), cognitive reappraisal ( OR=0.88, 95%CI =0.83-0.92), expressive suppression ( OR=1.08, 95%CI =1.02-1.15), and perceived campus atmosphere ( OR=0.97, 95%CI =0.95-0.98) were all related factors of psychological crisis among college students ( P <0.05). The qualitative analysis results showed that there were three themes for the influencing factors of college students psychological crisis, including differential impact of emotion regulation strategies on psychological state, shaping of psychological state of college students by family and bidirectional effect of perceived campus atmosphere on psychological state. Mixed analysis results showed that the influencing factors of college students psychological crisis were consistent in terms of emotion regulation strategies, and were expansive in terms of parenting style and perceived campus atmosphere. Conclusion Schools and mental health service departments can reduce the risk of psychological crisis by optimizing cognitive reappraisal and reducing expressive suppression, improve the level of psychological crisis by strengthening positive family interaction and blocking negative parenting style, and maintain the mental health level of college students by building a supportive campus environment and alleviating high pressure.

ObjectiveTo screen for the patients with metabolic dysfunction-associated fatty liver disease （MAFLD） among the physical examination population， to observe the characteristics of MAFLD patients， and to compare the differences in lifestyle between the MAFLD population and the non-MAFLD population. MethodsA cross-sectional study was conducted among 6 206 individuals who underwent physical examination in a physical examination institution in Beijing from December 2015 to December 2019， and according to the new diagnostic criteria for MAFLD， the examination population was divided into MAFLD group and non-MAFLD group. Based on body mass index （BMI）， the MAFLD group was further divided into lean MAFLD group （BMI<24 kg/m²） and non-lean MAFLD group （BMI ≥24 kg/m²）. Related data were compared between groups， including demographic indicators， education level， work pressure， physical measurement indicators， and lifestyles such as sleep， diet， and exercise. The independent-samples t test was used for comparison of normally distributed continuous data between two groups， the Mann-Whitney U test was used for comparison of non-normally distributed continuous data between two groups， and the chi-square test was used for comparison of categorical data between groups. ResultsOf all individuals in this study， 1 926 （31.1%） had MAFLD and 4 280 （68.9%） did not have MAFLD. Compared with the non-MAFLD group， the MAFLD group had significantly higher age （Z=-14.459， P<0.001）， proportion of male patients （χ²=72.004， P<0.001）， work pressure （χ²=7.744， P=0.005）， body weight （Z=-43.508， P<0.001）， BMI （Z=-47.621， P<0.001）， waist circumference （Z=-48.515， P<0.001）， hip circumference （Z=-42.121， P<0.001）， and waist-hip ratio （Z=-43.535， P<0.001）， as well as a significantly lower education level （χ²=33.583， P<0.001）. In terms of behavior， the MAFLD group had a significantly shorter sleep time （χ²=5.820， P=0.016） and a significantly faster eating speed （χ²=74.476， P<0.001）. In terms of diet， the patients in the MAFLD group consumed more high-sodium， high-sugar， and high-calorie diets （χ²=42.667， P<0.001） and low-fiber diet （χ²=4.367， P=0.008）. In terms of exercise， the MAFLD group had a significantly higher proportion of patients without exercise habits （χ²=10.278， P=0.001）. Further analysis showed that there were 202 individuals （10.5%） in the lean MAFLD group and 1 724 （89.5%） in the non-lean MAFLD group. Compared with the non-lean MAFLD group， the lean MAFLD group had significantly higher age （Z=3.368， P=0.001） and education level （χ²=9.647， P=0.002） and significantly lower proportion of male patients （χ²=27.664， P<0.001）， body weight （Z=-18.483， P<0.001）， BMI （Z=-23.286， P<0.001）， waist circumference （Z=-18.565， P<0.001）， and hip circumference （Z=-18.097， P<0.001）， and in terms of behavior， the non-lean MAFLD group had a significantly faster eating speed （χ²=4.549， P=0.033）. ConclusionThere is a relatively high prevalence rate of MAFLD among the physical examination population in Beijing， with a higher number of people with unhealthy lifestyles compared with the non-MAFLD population.

OBJECTIVES: To investigate the association of HOTAIR gene rs920778 single nucleotide polymorphism (SNP) with breast cancer susceptibility and response to HER2-targeted therapy in a Chinese population. METHODS: TaqMan probe-based real-time quantitative PCR was used for genotyping of the rs920778 locus (chr12:54,376,218) in peripheral blood genomic DNA from 287 breast cancer patients and 260 healthy individuals from northern Anhui Province. The genotype (GG, GT and TT) and allele (G/T) distribution frequencies were compared between the two groups to evaluate their association with breast cancer risk. Multivariate logistic regression analysis was conducted to assess the relationship between SNP at this locus and aggressive clinicopathological features (including tumor size, lymph node metastasis, ER/PR/HER2 status, and molecular subtypes) of breast cancer. For the HER2-positive subgroup, the association between rs920778 genotype and responses to dual-targeted therapy (trastuzumab [6 mg/kg q3w]+pertuzumab [420 mg q3w] + docetaxel [75 mg/m²]) was analyzed. The primary endpoints included pathological complete response rate (pCR), objective response rate (ORR), and progression-free survival (PFS). RESULTS: The TT genotype of rs920778 was associated with a significantly increased breast cancer susceptibility (OR=1.54, 95% CI: 1.09-2.19; P=0.017), an advanced tumor stage (P<0.001), lymph node metastasis (P<0.001), and the triple-negative subtype (P<0.001). In HER2-positive patients, TT genotype carriers had a markedly reduced objective response rate to dual HER2-targeted therapy (33.3% vs 89.3%, P=0.001) and a lower pathological complete response rate after neoadjuvant therapy (P=0.018). CONCLUSIONS The TT genotype of HOTAIR rs920778 serves as an independent risk factor for breast cancer susceptibility and aggressive progression in Chinese population and may predict the resistance to HER2-targeted therapies, suggesting its potential as a prognostic biomarker for precision oncology.
Adult ; Aged ; Female ; Humans ; Middle Aged ; Breast Neoplasms/drug therapy* ; Case-Control Studies ; China ; Drug Resistance, Neoplasm/genetics* ; Genetic Predisposition to Disease ; Genotype ; Polymorphism, Single Nucleotide ; Receptor, ErbB-2 ; RNA, Long Noncoding/genetics* ; East Asian People/genetics*

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.