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: 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 explore the effects of small-group collaborative stratified teaching in critical care medicine training for professional postgraduate students.Methods:We randomly assigned 71 professional postgraduate students who entered the Intensive Care Unit of The Affiliated Hospital of Qingdao University for standardized training between June 2020 and November 2020 into experimental group and control group. An entrance examination was taken after one week of unified training. Then the experimental group adopted small-group collaborative stratified teaching, while the control group adopted traditional teaching for training. After two months of training, the Mini-Clinical Evaluation Exercise (Mini-CEX) assessment, post competency assessment, exit examination, and teaching satisfaction evaluation were conducted. SPSS 25.0 was used for the t test and chi-square test. Results:In the Mini-CEX assessment, the experimental group had significantly higher scores in history-taking skills [(7.42±0.60) vs. (7.00±0.55)], physical examination [(7.47±0.56) vs. (6.94±0.24)], communication skills [(7.56±0.50）vs.(7.24±0.49)], clinical dialectical thinking [(7.53±0.56) vs. (7.03±0.39)], clinical judgement [(7.50±0.51) vs.(6.90±0.42)], organization/efficiency [(7.58±0.50) vs. (7.15±0.44)], and overall clinical competence [(7.64±0.49) vs. (7.17±0.39); all P<0.05] than the control group. In the post competency assessment, the experimental group had significantly better performance in clinical basic competence [(89.15±9.12) vs. (86.24±10.23)], medical knowledge application [(48.37±5.87) vs. (46.98±3.68)], teamwork [(48.10±3.55) vs. (45.96±4.83)], information and management [(68.52±7.61) vs. (66.38±5.54)], and academic research [(22.18±0.95) vs. (20.87±1.22); all P<0.05] than the control group. The experimental group was also significantly superior to the control group in terms of the exit examination score and teaching satisfaction (both P<0.05). Conclusions:Small-group collaborative stratified teaching can improve the quality of critical care medicine training for professional postgraduate students, and strengthen their clinical comprehensive abilities and post competencies.

Vascular calcification is a highly regulated process of ectopic calcification in cardiovascular system while no effective intervention can be clinically performed up to date.As vascular calcification undergoes a common regulatory mechanism within bone formation,bone morphogenetic protein 7(BMP-7)main-tains contractile phenotype of vascular smooth muscle cells and further inhibits vascular calcification via promoting the process of osteoblast differentiation,reducing ectopic calcification pressure by increasing bone formation and reducing bone resorption.This work systematically reviews the role of BMP-7 in vascular calcifi-cation and the possible mechanism,and their current clinical application as well.The current proceedings may help develope early diagnostic strategy and therapeutic treatment with BMP-7 as a new molecular marker and potential drug target.The expec-tation could achieve early prevention and intervention of vascular calcification and improve poor prognosis on patients.

Major depressive disorder(MDD)is a prevalent con-dition associated with substantial impairment and low remission rates.Traditional antidepressants demonstrate delayed effects,low cure rate,and inadequate therapeutic effectiveness for man-aging treatment-resistant depression(TRD).Several studies have shown that ketamine,a non-selective N-methyl-D-aspartate receptor(NMDAR)antagonist,can produce rapid and sustained antidepressant effects.Ketamine has demonstrated efficacy for reducing suicidality in TRD patients.However,the pharmaco-logical mechanism for ketamine's antidepressant effects remains incompletely understood.Previous research suggests that the an-tidepressant effects of ketamine may involve the monoaminergic,glutamatergic and dopaminergic systems.This paper provides an overview of the pharmacological mechanism for ketamine's anti-depressant effects and discuss the potential directions for future research.

Glutamate,norepinephrine,and their receptors com-prise the glutamatergic and norepinephrine systems,which mu-tually affect each other and play essential roles in mediating vari-ous neuropsychiatric diseases.This paper reviews the functions of N-methyl-D-aspartate receptor(NMDA-R)and α2-adrenergic receptor(α2-AR)and their functional crosstalk at the molecular level in brain in common neuropsychiatric diseases,which would benefit our understanding of neuropathophysiology of psychiatric diseases,drug development and optimization of clinical neuro-psychopharmacology.

Over the past 30 years,nano-drug delivery systems(NDDS)have become a promising field of drug research.However,a poor knowledge of the in vivo process of NDDS,the limited methods of pharmacokinetic correlation,and the inability to effectively provide strong support for the construction of upstream drug as well as the evaluation of downstream pharmacology and toxicology have become the technical bottleneck for their clinical transformation.Lipid nanodrug(LND)is the most successful NDDS for industrial transformation with great biocompatibility.Taking LND as an example,this paper reviewed the delivery process and influencing factors in vivo,and summarized the regulatory mecha-nism of biological environments on drug release in vivo.Based on advanced spectroscopy and mass spectrometry techniques,the spatial and temporal distribution of the dynamic carrier particle/depolymer-ized molecule ratio and dynamic free/encapsulated drug ratio of LND in biological matrix were ana-lyzed.Finally,the existing problems and future developments in this field were summarized to provide references for the analysis of NDDS in vivo,and stimulate readers'interest in nanomedical research and development.