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 To analyze the characteristics of pediatric medicines with priority review and approval for marketing in China， providing a reference for promoting enterprise R&D and production， as well as improving the supply guarantee mechanism for pediatric medicines. METHODS Based on publicly available data sources such as List of Approved Information for Pediatric Medications Subject to Priority Review and Approval， Pharnexcloud biomedical database， and National Medical Insurance Drug Directory， this study conducted a comprehensive analysis of the main characteristics of pediatric medicines with priority review and approval for marketing. RESULTS As of June 30， 2024， a total of 68 pediatric medicines had been approved through the priority review and approval process， covering 12 therapeutic areas， with oral dosage forms accounting for 64.71%. The median time from application to inclusion in priority review was 35.50 days， with an average of 41.69 days. The median time from inclusion in priority review to market approval was 1.24 years， with an average of 1.42 years. This included 12 domestic new medicines， 21 domestic generic medicines， 35 imported medicines， as well as 29 pediatric-specific medicines and 21 orphan medicines. Additionally， 31 of these medicines had been included in the medical insurance catalog， representing a proportion of 45.59%. CONCLUSIONS Currently， a trend of differentiated competition is emerging between domestic and imported pediatric medicines. The therapeutic areas for pediatric medicines are continuously expanding， and the dosage forms are becoming more tailored to children’s needs. However， there are still issues such as slow progress in new medicine development， insufficient stability in the medicine review and approval process， and a need to increase the proportion of medicines included in medical insurance.

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.

OBJECTIVE To investigate the effect and mechanism of Jingangteng capsules in the treatment of non-alcoholic fatty liver disease （NAFLD）. METHODS Thirty-two SD rats were randomly divided into normal group and modeling group. The modeling group was fed a high-fat diet to establish a NAFLD model. The successfully modeled rats were then randomly divided into model group， atorvastatin group[positive control， 2 mg/（kg·d）]， and Jingangteng capsules low- and high-dose groups [0.63 and 2.52 mg/（kg·d）]， with 6 rats in each group. The pathological changes of the liver were observed by hematoxylin-eosin staining and oil red O staining. Enzyme-linked immunosorbent assay was performed to determine the serum levels of triglycerides （TG）， total cholesterol （TC）， high-density lipoprotein cholesterol （HDL-C）， low-density lipoprotein cholesterol （LDL-C）， alanine transaminase （ALT）， aspartate transaminase （AST）， tumour necrosis factor-α （TNF-α）， interleukin-1β （IL-1β）， IL-6， IL-18. 16S rDNA amplicon sequencing and metabolomics techniques were applied to explore the effects of Jingangteng capsules on gut microbiota and metabolisms in NAFLD rats. Based on the E-mail：591146765@qq.com metabolomics results， Western blot analysis was performed to detect proteins related to the nuclear factor kappa-B （NF-κB）/NOD-like receptor family protein 3 （NLRP3） signaling pathway in the livers of NAFLD rats. RESULTS The experimental results showed that Jingangteng capsules could significantly reduce the serum levels of TG， TC， LDL-C， AST， ALT， TNF-α， IL-1β， IL-6， IL-18， while increased the level of HDL-C， and alleviated the hepatic cellular steatosis and inflammatory infiltration in NAFLD rats. They could regulate the gut microbiota disorders in NAFLD rats， significantly increased the relative abundance of Romboutsia and Oscillospira， and significantly decreased the relative abundance of Blautia （P＜0.05）. They also regulated metabolic disorders primarily by affecting secondary bile acid biosynthesis， fatty acid degradation， O-antigen nucleotide sugar biosynthesis， etc. Results of Western blot assay showed that they significantly reduced the phosphorylation levels of NF-κB p65 and NF-κB inhibitor α， and the protein expression levels of NLRP3， caspase-1 and ASC （P＜0.05 or P＜0.01）. CONCLUSIONS Jingangteng capsules could improve inflammation， lipid accumulation and liver injury in NAFLD rats， regulate the disorders of gut microbiota and metabolisms， and inhibit NF-κB/NLRP3 signaling pathway. Their therapeutic effects against NAFLD are mediated through the inhibition of the NF-κB/NLRP3 signaling pathway.

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.

ObjectiveTo investigate the pharmacodynamic effects of Houshihei San （HSHS） recorded with the effects of treating wind and limb heaviness on muscle tissue injury after middle cerebral artery occlusion （MCAO） in rats through the ferroptosis pathway. MethodsThirty SD male rats were selected and randomly grouped as follows： sham， MCAO， deferoxamine mesylate， high-dose HSHS （HSHS-H， 0.54 g·kg^-1）， and low-dose HSHS （HSHS-L， 0.27 g·kg^-1）， with 6 rats in each group. A laser scattering system was used to evaluate the stability of the MCAO model， and rats were administrated with corresponding agents by gavage for 7 days. During the administration period， behavioral， imaging and other methods were used to systematically evaluate the skeletal muscle tissue injury after MCAO and the therapeutic effect in each administration group. Hematoxylin-eosin staining was employed to evaluate the cross-section of muscle cells. Subsequently， immunohistochemistry was used to detect tumor suppressor p53 and glutathione peroxidase 4 （GPX4） in the soleus tissue. Western blot was employed to determine the protein levels of p53， GPX4， myogenic differentiation 1 （MyoD1）， nuclear factor E₂-related factor 2 （Nrf2）， Myostatin， solute carrier family 7 member 11 （SLC7A11）， muscle ring-finger protein-1 （MuRF1）， and muscle atrophy F-box protein （MAFbx） to verify the therapeutic effect in each group. ResultsCompared with the MCAO group， HSHS enhanced the locomotor ability and promoted muscle regeneration， which suggested that the pharmacological effects of HSHS were related to the inhibition of muscle tissue ferroptosis to reduce the expression of muscle atrophy factors. Behavioral and imaging results suggested that compared with the MCAO group， HSHS ameliorated neurological impairments in rats on day 7 （P<0.01）， enhanced 5-min locomotor distance and postural control （P<0.01）， strengthened grasping power and promoted muscle growth （P<0.01）， stabilized skeletal muscle length and weight （P<0.01）， and increased the cross-section of muscle cells （P<0.01）. Compared with the MCAO group， HSHS promoted the increases in glutathione and superoxide dismutase content and inhibited the increase in malondialdehyde content （P<0.05，P<0.01）. Ferroptosis pathway-related assays suggested that HSHS reduced the p53-positive cells and increased the GPX4-positive cells （P<0.01）. HSHS ameliorated muscle function decline after stroke by promoting the expression of GPX4， Nrf2， SLC7A11， and MyoD1 and inhibiting the expression of p53， Myostatin， MurRF1， and MAFbx to reduce ferroptosis in the muscle （P<0.01）. ConclusionHSHS， prepared with reference to the method in the Synopsis of Golden Chamber， can simultaneously reduce the myolysis and increase the protein synthesis in the skeletal muscle tissue after ischemic stroke by regulating the ferroptosis pathway.

Implementation Science is an interdisciplinary field dedicated to systematically studying how to effectively translate evidence-based research findings into practical application and implementation. In the health-related context, it focuses on enhancing the efficiency and quality of healthcare services, thereby facilitating the transition from scientific evidence to real-world practice. This article elaborates on Theories, Models, and Frameworks (TMF) within health-related Implementation Science, clarifying their basic concepts and classifications, and discussing their roles in guiding implementation processes. Furthermore, it reviews and prospects current research from three aspects: the constituent elements of TMF, their practical applications, and future directions. Five representative frameworks are emphasized, including the Consolidated Framework for Implementation Research (CFIR), the Practical Robust Implementation and Sustainability Model (PRISM), the Exploration, Preparation, Implementation, Sustainment (EPIS)framework, the Behavior Change Wheel (BCW), and the Normalization Process Theory (NPT). Additionally, resources such as the Dissemination & Implementation Models Webtool and the T-CaST tool are introduced to assist researchers in selecting appropriate TMFs based on project-specific needs.

AIM:To observe the clinical efficacy of vitrectomy at different times for open ocular trauma and explore the timing of stage Ⅱ vitrectomy.METHODS: Retrospective case series study. A total of 60 cases(60 eyes)with open ocular trauma who visited our ophthalmology department from June 2022 to February 2023 were included. They were divided into treatment group A(interval ≤14 d)and treatment group B(interval >14 d)based on the interval between the stage Ⅰ emergency treatment surgery and the stage Ⅱ vitreoretinal surgery. Among the 32 cases(32 eyes)in the treatment group A, 16 eyes(50%)had eyeball rupture, 13 eyes(41%)had penetrating injury, and 3 eyes(9%)had perforating injury. Among the 28 cases(28 eyes)in the treatment group B, 15 eyes(54%)had eyeball rupture, 12 eyes(43%)had penetrating injury, and one eye(4%)had perforating injury. The two groups of patients were followed-up for 6 mo after surgery, and the treatment effects were compared.RESULTS:There was no statistically significant difference in visual acuity between the two groups of patients before vitrectomy(P>0.05). In the treatment group A, 10 eyes(31%)had significantly improved visual acuity, 21 eyes(66%)had effectively enhanced visual acuity, and 1 eye(3%)had no improvement in visual acuity at 6 mo after surgery. Among the 28 eyes in the treatment group B, 5 eyes(18%)had significantly improved vision, 16 eyes(57%)had effectively enhanced vision, and 7 eyes(25%)had no change in vision, with statistically significant difference between the two groups(U=322.5, P=0.032). There was no significant difference between the treatment group A and the treatment group B in complications such as secondary glaucoma, silicone oil dependence, vitreous hemorrhage, and eyeball atrophy(P>0.05). There was no evidence of traumatic proliferative vitreoretinopathy(TPVR)in the treatment group A during postoperative follow-up, which was significantly lower than that of the treatment group B(P<0.05).CONCLUSION:The prognosis of the stage Ⅱ vitrectomy for open ocular injury is relatively good after completing the stage Ⅰ surgery within 2 wk.