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.

The rapid development of artificial intelligence (AI), especially generative AI (GenAI), has already brought, and will continue to bring, revolutionary changes to our daily production and life, as well as create new opportunities and challenges for diagnostic and therapeutic practices in the medical field. Haihe Hospital of Tianjin University collaborates with the National Supercomputer Center in Tianjin, Tianjin University, and other institutions to carry out research in areas such as smart healthcare, smart services, and smart management. We have conducted research and development of a full-process disease diagnosis and treatment assistance system based on GenAI in the field of smart healthcare. The development of this project is of great significance. The first goal is to upgrade and transform the hospital's information center, organically integrate it with existing information systems, and provide the necessary computing power storage support for intelligent services within the hospital. We have implemented the localized deployment of three models: Tianhe "Tianyuan", WiNGPT, and DeepSeek. The second is to create a digital avatar of the chief physician/chief physician's voice and image by integrating multimodal intelligent interaction technology. With generative intelligence as the core, this solution provides patients with a visual medical interaction solution. The third is to achieve deep adaptation between generative intelligence and the entire process of patient medical treatment. In this project, we have developed assistant tools such as intelligent inquiry, intelligent diagnosis and recognition, intelligent treatment plan generation, and intelligent assisted medical record generation to improve the safety, quality, and efficiency of the diagnosis and treatment process. This study introduces the content of a full-process disease diagnosis and treatment assistance system, aiming to provide references and insights for the digital transformation of the healthcare industry.
Artificial Intelligence ; Humans ; Delivery of Health Care ; Generative Artificial Intelligence

OBJECTIVE: This study examines utilizes the advantages of machine learning algorithms to discern key determinants in prognosticate postoperative circulatory complications (PCCs) for older patients. METHODS: This secondary analysis of data from a randomized controlled trial involved 1,720 elderly participants in five tertiary hospitals in Beijing, China. Participants aged 60-90 years undergoing major non-cardiac surgery under general anesthesia. The primary outcome metric of the study was the occurrence of PCCs, according to the European Society of Cardiology and the European Society of Anaesthesiology diagnostic criteria. The analysis metrics contained 67 candidate variables, including baseline characteristics, laboratory tests, and scale assessments. RESULTS: Our feature selection process identified key variables that significantly impact patient outcomes, including the duration of ICU stay, surgery, and anesthesia; APACHE-II score; intraoperative average heart rate and blood loss; cumulative opioid use during surgery; patient age; VAS-Move-Median score on the 1st to 3rd day; Charlson comorbidity score; volumes of intraoperative plasma, crystalloid, and colloid fluids; cumulative red blood cell transfusion during surgery; and endotracheal intubation duration. Notably, our Random Forest model demonstrated exceptional performance with an accuracy of 0.9872. CONCLUSION We have developed and validated an algorithm for predicting PCCs in elderly patients by identifying key risk factors.
Aged ; Aged, 80 and over ; Female ; Humans ; Male ; Middle Aged ; Cardiovascular Diseases/etiology* ; Machine Learning ; Postoperative Complications/etiology* ; Risk Factors ; Randomized Controlled Trials as Topic ; Secondary Data Analysis

OBJECTIVE: To investigate the association between long-term glycemic control and cerebral infarction risk in patients with diabetes through a large-scale cohort study. METHODS: This prospective, community-based cohort study included 12,054 patients with diabetes. From 2006 to 2012, 38,272 fasting blood glucose (FBG) measurements were obtained from these participants. FBG trajectory patterns were generated using latent mixture modelling. Cox proportional hazards models were applied to assess the subsequent risk of cerebral infarction associated with different FBG trajectory patterns. RESULTS: At baseline, the mean age of the participants was 55.2 years. Four distinct FBG trajectories were identified based on FBG concentrations and their changes over the 6-year follow-up period. After a median follow-up of 6.9 years, 786 cerebral infarction events were recorded. Different trajectory patterns were associated with significantly varied outcome risks (Log-Rank P < 0.001). Compared with the low-stability group, Hazard Ratio ( HR) adjusted for potential confounders were 1.37 for the moderate-increasing group, 1.23 for the elevated-decreasing group, and 2.08 for the elevated-stable group. CONCLUSION Sustained high FBG levels were found to play a critical role in the development of ischemic stroke among patients with diabetes. Controlling FBG levels may reduce the risk of cerebral infarction.
Humans ; Cerebral Infarction/blood* ; Middle Aged ; Male ; Female ; Blood Glucose/analysis* ; Fasting/blood* ; Aged ; Prospective Studies ; Risk Factors ; Diabetes Mellitus/blood* ; Adult ; Proportional Hazards Models

ObjectiveTo study the correlations between traditional Chinese medicine （TCM） syndromes and lipid metabolism in children with Wilson disease （WD）. MethodsClinical data and lipid metabolism indicators ［total cholesterol （TC）， triglycerides （TG）， low-density lipoprotein cholesterol （LDL-C）， high-density lipoprotein cholesterol （HDL-C）， apolipoprotein A1 （ApoA1）， apolipoprotein B （ApoB）， and lipoprotein a （Lpa）］ were retrospectively collected from 341 children with WD. The clinical data were compared among WD children with different syndromes， and the correlations between TCM syndromes and lipid metabolism in children with WD were analyzed. Least absolute shrinkage and selection operator （LASSO） regression was used for variable screening， and unordered multinomial Logistic regression was employed to analyze the effects of lipid metabolism indicators on TCM syndromes. ResultsThe 341 children with WD included 121 （35.5%） children with the dampness-heat accumulation syndrome， 103 （30.2%） children with the liver-kidney Yin deficiency syndrome， 68 children with the combined phlegm and stasis syndrome， 29 children with the spleen-kidney Yang deficiency syndrome， and 20 children with the liver qi stagnation syndrome. The liver-kidney Yin deficiency syndrome， combined phlegm and stasis syndrome， and spleen-kidney Yang deficiency syndrome had correlations with the levels of lipid metabolism indicators （P<0.05）. Lipid metabolism abnormalities occurred in 232 （68.0%） children， including hypertriglyceridemia （108）， hypercholesterolemia （23）， mixed hyperlipidemia （67）， lipoprotein a-hyperlipoproteinemia （12）， and hypo-HDL-cholesterolemia （22）. The percentages of hypertriglyceridemia and hypo-HDL-cholesterolemia varied among children with different TCM syndromes （P<0.05）. Correlations existed for the liver-kidney Yin deficiency syndrome with TG， TC， and HDL-C， the combined phlegm and stasis syndrome with TG， the spleen-kidney Yang deficiency syndrome with TG， TC， and LDL-C， and the liver Qi stagnation syndrome with TC and LDL-C （P<0.05， P<0.01）. ConclusionThe TCM syndromes of children with WD are dominated by the dampness-heat accumulation syndrome and the liver-kidney Yin deficiency syndrome， and dyslipidemia in the children with WD is dominated by hypertriglyceridemia and mixed hyperlipidemia. There are different correlations between TCM syndromes and lipid metabolism indicators， among which TG， TC， LDL-C， and HDL-C could assist in identifying TCM syndromes in children with WD.

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.

Cardiac fibrosis is characterized by an elevated amount of extracellular matrix (ECM) within the heart. However, the persistence of cardiac fibrosis ultimately diminishes contractility and precipitates cardiac dysfunction. Circular RNAs (circRNAs) are emerging as important regulators of cardiac fibrosis. Here, we elucidate the functional role of a specific circular RNA CELF1 in cardiac fibrosis and delineate a novel feedback loop mechanism. Functionally, circ-CELF1 was involved in enhancing fibrosis-related markers' expression and promoting the proliferation of cardiac fibroblasts (CFs), thereby exacerbating cardiac fibrosis. Mechanistically, circ-CELF1 reduced the ubiquitination-degradation rate of BRPF3, leading to an elevation of BRPF3 protein levels. Additionally, BRPF3 acted as a modular scaffold for the recruitment of histone acetyltransferase KAT7 to facilitate the induction of H3K14 acetylation within the promoters of the Celf1 gene. Thus, the transcription of Celf1 was dramatically activated, thereby inhibiting the subsequent response of their downstream target gene Smad7 expression to promote cardiac fibrosis. Moreover, Celf1 further promoted Celf1 pre-mRNA transcription and back-splicing, thereby establishing a feedback loop for circ-CELF1 production. Consequently, a novel feedback loop involving CELF1/circ-CELF1/BRPF3/KAT7 was established, suggesting that circ-CELF1 may serve as a potential novel therapeutic target for cardiac fibrosis.