Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

BACKGROUND:We have successfully established an animal model of small ear pig in southern Yunnan province with internal tension relieving technique combined with autologous Achilles tendon for anterior cruciate ligament reconstruction,and verified the stability and reliability of the model.However,whether internal tension relieving technique can promote the ligamentalization process of autologous Achilles tendon graft has not been studied. OBJECTIVE:To investigate the differences in the process of ligamentalization between conventional reconstruction and internal reduction reconstruction of the anterior cruciate ligament by gross view,histology and electron microscopy. METHODS:Thirty adult female small ear pigs in southern Yunnan province were selected.Anterior cruciate ligament reconstruction was performed on the left knee joint with the ipsilateral knee Achilles tendon(n=30 in the normal group),and anterior cruciate ligament reconstruction was performed on the right knee joint with the ipsilateral knee Achilles tendon combined with the internal relaxation and enhancement system(n=30 in the relaxation group).The autogenous right forelimb was used as the control group;the anterior cruciate ligament was exposed but not severed or surgically treated.At 12,24,and 48 weeks after surgery,10 animals were sacrificed,respectively.The left and right knee joint specimens were taken for gross morphological observation to evaluate the graft morphology.MAS score was used to evaluate the excellent and good rate of the ligament at each time point.Hematoxylin-eosin staining was used to evaluate the degree of ligament graft vascularization.Collagen fibers and nuclear morphology were observed,and nuclear morphology was scored.Ultrastructural remodeling was evaluated by scanning electron microscopy and transmission electron microscopy. RESULTS AND CONCLUSION:(1)The ligament healing shape of the relaxation group was better at various time points after surgery,and the excellent and good rate of MAS score was higher(P<0.05).Moreover,the relaxation group could obtain higher ligament vascularization score(P<0.05).(2)The arrangement of collagen bundles and fiber bundles in the two groups gradually tended to be orderly,and the transverse fiber connections between collagen gradually increased and thickened,suggesting that the strength and shape degree of the grafts were gradually improved,but the ligament remodeling in the relaxation group was always faster than that in the normal group at various time points after surgery.(3)The diameter,distribution density,and arrangement degree of collagen fibers in the relaxation group were better than those in the normal group at all time points,especially in the comparison of collagen fiber diameter between and within the relaxation group(P<0.05).

Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

As a hot spot in clinical research today, immune checkpoint inhibitor has been recommended by guidelines in the first- and second-line treatments of advanced cervical cancer as immune monotherapy or combination therapy. It has also achieved good efficacy in clinical practice. In locally advanced cervical cancer, immune checkpoint inhibitors have been included in the guidelines for adjuvant therapy, and good tumor regression effects have been achieved in clinical practice. Based on the results of existing trials, immune checkpoint inhibitors have also shown good clinical potential as neoadjuvant therapy. Furthermore, the issue of immunotherapy rechallenge has increasingly captured clinicians’ attention, offering a potential new therapeutic strategy for cervical cancer patients with prior immunotherapy exposure. In this article, the clinical application and research progress of immune checkpoint inhibitors in the treatment of cervical cancer in recent years are summarized to provide valuable ideas and directions for clinical treatment.

Objective: To investigate the vitamin D levels in children aged 1 year in Shaoxing City, Zhejiang Province, so as to provide the basis for prevention and treatment of vitamin D deficiency in children and promoting their health. Methods: The 1-year-old children who underwent health examinations at the Department of Child Health Care of Shaoxing Maternal and Child Health Care Hospital from September 2023 to August 2024 were selected. Basic information of the children was collected through the medical record information system, and their length and weight were measured. The length, weight and nutritional status were evaluated according to the Technical Specifications for the Management of Nutritional Diseases in Children. Serum 25-hydroxyvitamin D [25- (OH) D] levels were measured using electrochemiluminescence assay, and vitamin D levels were assessed based on the fifth edition of Child Health Care. The vitamin D levels were analyzed among the children with different genders, testing months, and growth status. Results: A total of 2 245 children were recruited, including 1 189 boys (52.96%) and 1 056 girls (47.04%). The median serum 25- (OH) D level was 39.98 (interquartile range, 16.63) ng/mL. Vitamin D insufficiency was observed in 279 children, with an insufficiency rate of 12.43%. The median serum 25- (OH) D level in boys was 39.26 (interquartile range, 17.75) ng/mL, which was lower than that in girls at 41.39 (17.75) ng/mL (P<0.05). The vitamin D insufficiency rate was 13.04% in boys and 11.74% in girls, with no statistically significant difference (P>0.05). The lowest vitamin D insufficiency rate was observed in August at 4.95%, while the highest rate was in September at 23.89%, showing the statistically significant difference across testing months (P<0.05). The children with above-average length ratings, higher weight ratings and obesity had higher vitamin D insufficiency rates, at 17.29%, 20.86% and 20.88%, respectively. The vitamin D insufficiency rate increased with higher weight and nutritional status ratings (both P<0.05), but no significant change was observed with higher length ratings (P>0.05). Conclusions The vitamin D insufficiency rate among 1-year-old children in Shaoxing City was 12.43%, with variations observed in different testing months, weight and nutritional status. Targeted prevention and intervention measures should be implemented to address these differences.

Objective To investigate the levels of gross α and gross β activities in different water types within a 40-kilometer radius around the Zhangzhou Nuclear Power Plant prior to its operation. Methods In 2018, drinking water samples were collected from the area surrounding the nuclear power plant during both the wet and dry seasons, including source water, treated water, tap water, and well water. The gross α and gross β activity concentrations were measured using a low-background α/β counter, followed by statistical analysis. Results A total of 80 water samples from different sources around the Zhangzhou Nuclear Power Plant were collected. The average gross α and gross β activity concentrations during the wet season were (0.110 ± 0.036) Bq/L and (0.643 ± 0.028) Bq/L, respectively, while those during the dry season were (0.124 ± 0.032) Bq/L and (0.624 ± 0.026) Bq/L, respectively. There were no significant differences in the gross α and gross β activity concentrations between the wet and dry seasons for the overall sample set (P > 0.05). However, there were statistically significant differences in the gross α and gross β activity concentrations between the wet and dry seasons for source water and well water (Z_wet = −2.005, −2.123; Z_dry = −1.943, −3.090; P < 0.05). Conclusion The radioactivity levels in different water types within various ranges around the Zhangzhou Nuclear Power Plant before its operation were determined. The measured activity concentrations were at the same level as those from previous investigations in other regions of Fujian Province.

Objective To compare the cost and utility of aflibercept and conbercept for the treatment of wet age-related macular degeneration(wetAMD),in order to provide a reference for the selection of treatment regimens from the perspective of pharmacoeconomics.Methods The incremental cost-utility ratio(ICUR)was obtained by using Markov model to simulate the survival of the two treatment regimens over the 5-year period,calculating costs and health outputs separately.Univariate sensitivity analysis was used to determine the impact of the parameter on ICUR,and probability sensitivity analysis was used to determine the influence of the uncertainty of each model parameter on the research results.One times the 2022 gross domestic product(GDP)per capita of China was used as the willingness-to-pay threshold(WTP)to judge its economy.Results Over the simulation period,the compazine regimen was significantly economical against the aflibercept regimen,with an ICUR of-1 528 840 per quality-adjusted life year(QALY),which was lower than the WTP.Univariate sensitivity analysis showed that the transition probability between mild and moderate visual status between the two regimens and the number of aflibercept injections per year were significant influencing factors of ICUR.Probabilistic sensitivity analysis pointed to a significant cost-utility advantage for conbercept at a WTP of one times GDP(probability of 65.9％),which was a more robust result.Conclusion For the treatment of wetAMD,conbercept has a cost-utility advantage compared with aflibercept.