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 molecular mechanism of chronic osteomyelitis and to clarify the role of MAPK signal pathway in the pathogenesis of chronic osteomyelitis,by collecting and analyzing the transcriptional information of bone tissue in patients with chronic osteomyelitis.Methods Four cases of traumatic osteomyelitis in limbs from June 2019 to June 2020 were selected,and the samples of necrotic osteonecrosis from chronic osteomyelitis(necrotic group),and normal bone tissue(control group)were collected.Transcriptome information was collected by Illumina Hiseq Xten high throughput sequencing platform,and the gene expression in bone tissue was calculated by FPKM.The differentially expressed genes were screened by comparing the transcripts of the Necrotic group and control group.Genes were enriched by GO and KEGG.MAP3K7 and NFATC1 were selected as differential targets in the verification experiments,by using rat osteomyelitis animal model and im-munohistochemical analysis.Results A total of 5548 differentially expressed genes were obtained by high throughput sequenc-ing by comparing the necrotic group and control group,including 2701 up-regulated and 2847 down-regulated genes.The genes enriched in MAPK pathway and osteoclast differentiation pathway were screened,the common genes expressed in both MAPK and osteoclast differentiation pathway were(inhibitor of nuclear factor κ subunit Beta,IκBKβ),(mitogen-activated protein ki-nase 7,MAP3K7),(nuclear factor of activated t cells 1,NFATC1)and(nuclear factor Kappa B subunit 2,NFκB2).In rat os-teomyelitis model,MAP3K7 and NFATC1 were highly expressed in bone marrow and injured bone tissue.Conclusion Based on the transcriptome analysis,the MAPK signaling and osteoclast differentiation pathways were closely related to chronic os-teomyelitis,and the key genes IκBKβ,MAP3K7,NFATC1,NFκB2 might be new targets for clinical diagnosis and therapy of chronic osteomyelitis.

Objective To investigate the expression and prognostic value of spindle and kinetochore-associated complex subunit 2(SKA2)in cervical cancer tissues,as well as its impact on the proliferation,migration and invasion of cervical cancer cells.Methods The expression of SKA2 in cervical cancer tissues was analyzed by bioinformatics database and immunohistochemical SP method,and the relationship between SKA2 expression level and clinicopathological features of cervical cancer patients and its prognostic value was analyzed.The mRNA expression of SKA2 in human normal cervical cells(HcerEpic)and cervical cancer cells(HeLa,SiHa,CaSki,C-33A)was detected by RT-qPCR.Cervical cancer cells SiHa with higher SKA2 expression level was selected for further study.SiHa cell model with down-regulated SKA2 expression was constructed,and its knockdown effect was verified.Cell proliferation capacity was detected by CCK-8 method,cell migration capacity was detected by cell scratch wound healing assay,and cell migration and invasion capacity was detected by Transwell assay.Results Compared with normal cervical tissues and cells,the expression levels of SKA2 mRNA and protein were higher in cervical cancer tissues and cells,and the differences were statistically significant(P<0.05).High SKA2 expression was associated with FIGO staging in patients with cervical cancer.Furthermore,SKA2 knockdown could inhibit the proliferation,migration and invasion of SiHa cells in cervical cancer(P<0.05).Conclusion SKA2 is up-regulated in cervical cancer tissues and cells,and can promote the proliferation,migration and invasion of cervical cancer cells.The expression level of SKA2 is associated with the progression of cervical cancer,and the prognosis of cervical cancer patients with high SKA2 expression is worse.

Objective To compare the efficacy of renal proximal renal artery denervation(pRDN)and full-length renal artery denervation(fRDN)for treatment of hypertension.Methods Fifty-six hypertensive patients were enrolled and randomly assigned to full-length renal artery denervation group(n=25)and proximal renal artery denervation group(n=31).After the procedure,24-hour ambulatory blood pressure monitoring(24 h-ABPM)at 6 months and office blood pressure at 12 months was recorded for statistical analysis.Results The blood pressure at follow-up reduced significantly in both groups,while there was no significant difference between groups.The baseline office blood pressure in fRDN group and pRDN group was(180±15)/(104±10)mmHg and(180±12)/(103±8)mmHg,respectively,which decreased to(142±9)/(82±7)mmHg and(143±10)/(83±6)mmHg at 12 months postoperatively(P＜0.001 within groups and P＞0.05 between groups).The baseline 24 h-ABPM in the two groups was(162±13)/(95±8)mmHg and(160±12)/(94±8)mmHg,respectively,which decreased to(142±11)/(83±7)mmHg and(141±8)/(81±7)mmHg at 6 months postoperatively(P＜0.001 within groups and P＞0.05 between groups).However,there was no significant difference in the reduction of office blood pressure and ambulatory blood pressure between the two groups.No treatment-related adverse events were observed.Conclusions pRDN has similar antihypertensive effect to fRDN.

Veno-arterial extracorporeal membrane oxygenation is an effective method to reduce perioperative adverse events such as cardiogenic shock in patients undergoing complex high-risk indicated percutaneous coronary intervention.Femoral artery and femoral vein are the main routes for conventional veno-arterial extracorporeal membrane oxygenation in China,while the cases of extracorporeal membrane oxygenation insertion via axillary artery are relatively rare.However,the axillary artery intubation veno-arterial extracorporeal membrane oxygenation assisted mode has been regarded as one of the routine clinical paths for the treatment of critically ill patients in foreign countries.This paper reports a case of an elderly male patient who underwent high risk and complex percutaneous coronary interventional therapy by right axillary artery implantation with extracorporeal membrane oxygenation assisted circulation due to the difficulty of femoral artery approach.In order to provide reference for the selection of clinical extracorporeal membrane oxygenation technique route.

Objective:To investigate the factors related to the tumor growth in subjects with low-risk papillary thyroid microcarcinoma (PTMC) based on ultrasonography.Methods:This was a cross-sectional study. A total of 136 subjects who received health check-up in Health Management Center, the Affiliated Zhongshan Hospital of Dalian University from October 2017 to December 2023 were enrolled in the study. Low-risk PTMC were detected by ultrasonogrphy in those subjects and ultrashonography was followed up to observe the changes of maximum diameter and volume of the tumor, and metastasis of cervical lymph nodes. The clinical characteristics and ultrasonic image features were compared between the subjects with the tumor growth and without tumor growth, and the influencing factors of tumor growth were analyzed.Results:Among 136 subjects with low-risk PTMC, there were 23 cases (16.9%) with tumor growth (growth group) and 113 cases (83.1%) without tumor growth (non-growth group). Cervical lymph node metastasis occurred in 8 cases (5.9%: 7 (30.4%) in the growth group and 1 (0.9%) in non-growth group), no distant metastasis were detected. There were significantly differences in patients age of initial diagnoisi, maximum diameter and volume of tumors between the growth group and non-growth group (all P<0.05). Logistic regression analysis showed that age of initial diagnoisi ≤40 years ( OR=4.299, 95% CI:1.662-12.175, P=0.003) was an independent risk factor for tumor growth and the maximum diameter of the initial examination was independent protective factor for tumor growth (increasing 1 mm of initial diameter: OR=0.554, 95% CI:0.317-0.969, P=0.038). Conclusion:The size of most low-risk PTMC detected by ultrasonography during the health check-up does not grow and the risk of cervical lymph node metastasis is low; however, for those with age of initial diagnoisi ≤40 years and smaller size tumor, the risk of PTMC growth would be increased.