An-Tian Chen
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
An-Tian Chen has a total of 1000 co-authors. Here are the top co-authors by number of shared publications:
- Wei Chen
Department of Cardiology, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
3 shared publications - Wenle Li
State Key Laboratory of Vaccines for Infectious Diseases Xiang An Biomedicine Laboratory National Innovation Platform for Industry‐Education Integration in Vaccine Research School of Public Health Xiamen University Xiamen China
3 shared publications Anfa Chen
Department of Orthopedics, Jiangxi Province Hospital of Integrated Chinese & Western Medicine, China
2 shared publicationsChan Xu
State Key Laboratory of Molecular Vaccinology and Molecular Diagnostics & Center for Molecular Imaging and Translational Medicine, School of Public Health, Xiamen University, Xiamen, China
2 shared publicationsChengliang Yin
Faculty of Medicine, Macau University of Science and Technology, Macau, China
2 shared publicationsHaizhen Ding
State Key Laboratory of Molecular Vaccinology and Molecular Diagnostics & Center for Molecular Imaging and Translational Medicine, School of Public Health, Xiamen University, Xiamen, China
2 shared publications
Latest Publications
- Open Access
- 1 January 2026
- Published by Walter de Gruyter GmbH in Asian Journal of Medical Humanities
- Vol. 5 (1)
The convergence of large language models and physician–patient relationship research: a bibliometric analysis
Objectives: This study systematically reviewed publications on natural language processing and large language models in the field of physician–patient relationships using bibliometric methods, and analyzed publication trends, current research status, and thematic evolution. Methods: Publications were selected from the PubMed database. Bibliometric research was conducted using CiteSpace, including keyword co-occurrence and clustering, timeline analysis, and keyword burst analysis. Results: A total of 128 publications were included. Literature fluctuated at a low level before 2019, but increased markedly following the release of ChatGPT, peaking in 2024. Keyword co-occurrence and clustering identified seven core thematic categories: artificial intelligence, natural language processing, large language models, patient education, clinical decision-making, social media, care experience, and e-Health. Timeline analysis indicated a recent shift toward shared decision-making and emotional support. Keyword burst analysis further revealed that large language models, clinical decision support, and generative AI emerged as prominent topics after 2024. Conclusions: The rise of large language models has accelerated research in the physician–patient relationship domain, shifting research hotspots toward human-centered scenarios, particularly in areas such as patient experience, clinical decision support, and e-Health.
- 1 January 2026
- Published by Walter de Gruyter GmbH in Asian Journal of Medical Humanities
- Vol. 5 (1)
Open AccessThe convergence of large language models and physician–patient relationship research: a bibliometric analysis
Objectives: This study systematically reviewed publications on natural language processing and large language models in the field of physician–patient relationships using bibliometric methods, and analyzed publication trends, current research status, and thematic evolution. Methods: Publications were selected from the PubMed database. Bibliometric research was conducted using CiteSpace, including keyword co-occurrence and clustering, timeline analysis, and keyword burst analysis. Results: A total of 128 publications were included. Literature fluctuated at a low level before 2019, but increased markedly following the release of ChatGPT, peaking in 2024. Keyword co-occurrence and clustering identified seven core thematic categories: artificial intelligence, natural language processing, large language models, patient education, clinical decision-making, social media, care experience, and e-Health. Timeline analysis indicated a recent shift toward shared decision-making and emotional support. Keyword burst analysis further revealed that large language models, clinical decision support, and generative AI emerged as prominent topics after 2024. Conclusions: The rise of large language models has accelerated research in the physician–patient relationship domain, shifting research hotspots toward human-centered scenarios, particularly in areas such as patient experience, clinical decision support, and e-Health.
- Open Access
- 5 November 2025
- Published by Wolters Kluwer Health in International Journal of Surgery
- Vol. 111 (12) , 10017-10029
Characteristics, cost/effect consideration of clinical examinations, and construction of machine learning models of restrictive cardiomyopathy: insights from Peking Union Medical College Hospital
Background: Restrictive cardiomyopathy (RCM) is an uncommon condition with heterogeneous causes. Amyloidosis, a major subtype, presents with diagnostic complexity, economic burden, and prognostic implications. This study aimed to apply machine learning (ML) techniques to improve the diagnosis of amyloidosis among RCM patients and assess the cost-effectiveness of laboratory tests. Methods: This study included patients with RCM who underwent transthoracic echocardiography (TTE) and cardiac magnetic resonance imaging (MRI). Feature selection was performed using the least absolute shrinkage and selection operator regression based on variables that showed statistical differences between groups. These selected features were then used to construct eight ML models, which were trained and validated using leave-one-out cross-validation. The best-performing model was evaluated for sample size and interpreted using Shapley additive explanations to enhance model transparency. Laboratory testing costs related to autoimmune, infection, tumor, and amyloidosis evaluations were compared across subgroups. Results: The Random Forest (RF) model achieved the best performance, with an area under the curve of 0.977, an accuracy of 0.908, a sensitivity of 0.869, and a specificity of 0.927. The model also showed a favorable Brier score and a satisfying effect size, indicating good performance in distinguishing amyloidosis from other RCM subtypes. Cost analysis revealed that patients without underlying autoimmune, infectious, or tumor-related etiologies incurred unnecessary expenditures. Multivariate regression identified key imaging features associated with amyloidosis, including left ventricular posterior wall thickness and left ventricular ejection fraction (LVEF) from TTE, and left ventricular short axis diameter, LVEF, and interventricular septum thickness from cardiac MRI. Conclusion: This study established an interpretable ML model based on the RF algorithm and accurately distinguished amyloidosis among RCM patients. By guiding more targeted use of amyloidosis-specific testing, the model offers a potential cost-saving strategy while improving diagnostic efficiency. These findings support the clinical integration of ML-based tools to streamline decision-making and optimize the allocation of healthcare resources.
- 5 November 2025
- Published by Wolters Kluwer Health in International Journal of Surgery
- Vol. 111 (12) , 10017-10029
Open AccessCharacteristics, cost/effect consideration of clinical examinations, and construction of machine learning models of restrictive cardiomyopathy: insights from Peking Union Medical College Hospital
Background: Restrictive cardiomyopathy (RCM) is an uncommon condition with heterogeneous causes. Amyloidosis, a major subtype, presents with diagnostic complexity, economic burden, and prognostic implications. This study aimed to apply machine learning (ML) techniques to improve the diagnosis of amyloidosis among RCM patients and assess the cost-effectiveness of laboratory tests. Methods: This study included patients with RCM who underwent transthoracic echocardiography (TTE) and cardiac magnetic resonance imaging (MRI). Feature selection was performed using the least absolute shrinkage and selection operator regression based on variables that showed statistical differences between groups. These selected features were then used to construct eight ML models, which were trained and validated using leave-one-out cross-validation. The best-performing model was evaluated for sample size and interpreted using Shapley additive explanations to enhance model transparency. Laboratory testing costs related to autoimmune, infection, tumor, and amyloidosis evaluations were compared across subgroups. Results: The Random Forest (RF) model achieved the best performance, with an area under the curve of 0.977, an accuracy of 0.908, a sensitivity of 0.869, and a specificity of 0.927. The model also showed a favorable Brier score and a satisfying effect size, indicating good performance in distinguishing amyloidosis from other RCM subtypes. Cost analysis revealed that patients without underlying autoimmune, infectious, or tumor-related etiologies incurred unnecessary expenditures. Multivariate regression identified key imaging features associated with amyloidosis, including left ventricular posterior wall thickness and left ventricular ejection fraction (LVEF) from TTE, and left ventricular short axis diameter, LVEF, and interventricular septum thickness from cardiac MRI. Conclusion: This study established an interpretable ML model based on the RF algorithm and accurately distinguished amyloidosis among RCM patients. By guiding more targeted use of amyloidosis-specific testing, the model offers a potential cost-saving strategy while improving diagnostic efficiency. These findings support the clinical integration of ML-based tools to streamline decision-making and optimize the allocation of healthcare resources.
- Open Access
- 11 July 2025
- Published by International Society of Global Health in Journal of Global Health
- Vol. 15, 04191
Exploring the complex relationship between systemic lupus erythematosus and coronavirus disease 2019: genetic insights and potential protective mechanisms
Coronavirus disease 2019 (COVID-19) presents a multifaceted threat to the human health, impacting not only the respiratory system but also triggering systemic inflammatory responses. These responses can escalate into a cytokine storm, overwhelming the body’s defences and leading to severe complications such as acute respiratory distress syndrome [1]. In systemic lupus erythematosus (SLE), immune dysregulation […]
- 11 July 2025
- Published by International Society of Global Health in Journal of Global Health
- Vol. 15, 04191
Open AccessExploring the complex relationship between systemic lupus erythematosus and coronavirus disease 2019: genetic insights and potential protective mechanisms
Coronavirus disease 2019 (COVID-19) presents a multifaceted threat to the human health, impacting not only the respiratory system but also triggering systemic inflammatory responses. These responses can escalate into a cytokine storm, overwhelming the body’s defences and leading to severe complications such as acute respiratory distress syndrome [1]. In systemic lupus erythematosus (SLE), immune dysregulation […]
- 19 October 2024
- Published by Elsevier in The Spine Journal
- Vol. 25 (3) , 460-473
Development of machine learning model for predicting prolonged operation time in lumbar stenosis undergoing posterior lumbar interbody fusion: a multicenter study
- 19 October 2024
- Published by Elsevier in The Spine Journal
- Vol. 25 (3) , 460-473
Development of machine learning model for predicting prolonged operation time in lumbar stenosis undergoing posterior lumbar interbody fusion: a multicenter study
- Open Access
- 19 September 2024
- Published by Oxford University Press (OUP) in ESC Heart Failure
- Vol. 12 (1) , 353-368
Explainable Machine Learning and Online Calculators to Predict Heart Failure Mortality in Intensive Care Units
Aims: This study aims to develop explainable machine learning models and clinical tools for predicting mortality in patients in the intensive care unit (ICU) with heart failure (HF). Methods: Patients diagnosed with HF who experienced their first ICU stay lasting between 24 h and 28 days were selected from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database. The primary outcome was all-cause mortality within 28 days. Data analysis was performed using Python and R, with feature selection conducted via least absolute shrinkage and selection operator (LASSO) regression. Fifteen models were evaluated, and the most effective model was rendered explainable through the Shapley additive explanations (SHAP) approach. A nomogram was developed based on logistic regression to facilitate interpretation. For external validation, the eICU database was utilized. Results: After selection, the study included 2343 records, with 1808 surviving and 535 deceased patients. The median age of the study population was 70.00, with ~3/5 males (60.31%). The median length of stay in the ICU was 6.00 days. The median age of the survival group was younger than the non-survival group (69.00 vs. 73.00), and non-survival patients spent longer time in the ICU. Seventy-five features were initially selected, including basic information, vital signs, laboratory tests, haemodynamics and oxygen status. LASSO regression determined the shrinkage parameter α = 0.020, and 44 features were chosen for model construction. The linear discriminant analysis (LDA) model showed the best performance, and the accuracy reached 0.8354 in the training cohort and 0.8563 in the testing cohort. It showed satisfying area under the curve (AUC), recall, precision, F1 score, Cohen’s kappa score and Matthew’s correlation coefficient. The concordance index (c-index) reached 0.7972 in the training cohort and 0.8125 in the testing cohort. In external validation, the LDA model achieved approximately 0.9 in accuracy, precision, recall and F1 score, with an AUC of 0.79. Univariable analysis was performed in the training cohort. Features that differed significantly between the survival and non-survival groups were subjected to multiple logistic regression. The nomogram built on multiple logistic regression included 14 features and demonstrated excellent performance. The AUC of the nomogram is 0.852 in the training cohort, 0.855 in the internal validation cohort and 0.770 in the external validation cohort. The calibration curve showed good consistency. Conclusions: The study developed an LDA and a nomogram model for predicting mortality in HF patients in the ICU. The SHAP approach was employed to elucidate the LDA model, enhancing its utility for clinicians. These models were made accessible online for clinical application.
- 19 September 2024
- Published by Oxford University Press (OUP) in ESC Heart Failure
- Vol. 12 (1) , 353-368
Open AccessExplainable Machine Learning and Online Calculators to Predict Heart Failure Mortality in Intensive Care Units
Aims: This study aims to develop explainable machine learning models and clinical tools for predicting mortality in patients in the intensive care unit (ICU) with heart failure (HF). Methods: Patients diagnosed with HF who experienced their first ICU stay lasting between 24 h and 28 days were selected from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database. The primary outcome was all-cause mortality within 28 days. Data analysis was performed using Python and R, with feature selection conducted via least absolute shrinkage and selection operator (LASSO) regression. Fifteen models were evaluated, and the most effective model was rendered explainable through the Shapley additive explanations (SHAP) approach. A nomogram was developed based on logistic regression to facilitate interpretation. For external validation, the eICU database was utilized. Results: After selection, the study included 2343 records, with 1808 surviving and 535 deceased patients. The median age of the study population was 70.00, with ~3/5 males (60.31%). The median length of stay in the ICU was 6.00 days. The median age of the survival group was younger than the non-survival group (69.00 vs. 73.00), and non-survival patients spent longer time in the ICU. Seventy-five features were initially selected, including basic information, vital signs, laboratory tests, haemodynamics and oxygen status. LASSO regression determined the shrinkage parameter α = 0.020, and 44 features were chosen for model construction. The linear discriminant analysis (LDA) model showed the best performance, and the accuracy reached 0.8354 in the training cohort and 0.8563 in the testing cohort. It showed satisfying area under the curve (AUC), recall, precision, F1 score, Cohen’s kappa score and Matthew’s correlation coefficient. The concordance index (c-index) reached 0.7972 in the training cohort and 0.8125 in the testing cohort. In external validation, the LDA model achieved approximately 0.9 in accuracy, precision, recall and F1 score, with an AUC of 0.79. Univariable analysis was performed in the training cohort. Features that differed significantly between the survival and non-survival groups were subjected to multiple logistic regression. The nomogram built on multiple logistic regression included 14 features and demonstrated excellent performance. The AUC of the nomogram is 0.852 in the training cohort, 0.855 in the internal validation cohort and 0.770 in the external validation cohort. The calibration curve showed good consistency. Conclusions: The study developed an LDA and a nomogram model for predicting mortality in HF patients in the ICU. The SHAP approach was employed to elucidate the LDA model, enhancing its utility for clinicians. These models were made accessible online for clinical application.
- 1 September 2024
- Published by Wolters Kluwer Health in Hypertension
- Vol. 81 (Suppl_1) , AP135
Abstract P135: Evaluating the Impact of Cumulative Blood Pressure and Metabolic Indicators on Long-Term Cardiovascular Risk: Insights from a 9-year Nationwide Prospective Cohort Study
Background: Cardiovascular diseases (CVD) are a global health challenge, particularly for older adults. Blood pressure and the Triglyceride-Glucose (TyG) index, an insulin resistance indicator, are key predictors of cardiometabolic risks. Methods: Using data from the China Health and Retirement Longitudinal Study (CHARLS), this study followed participants without baseline CVD over 9 years. We stratified subjects by cumulative TyG and TyG-BMI indices, applying Cox proportional hazards adjusted by age and gender, and restricted cubic spline regression for CVD risk evaluation. Results: From 11,847 participants, 4,751 were analyzed, mainly females (55.25%), with a median age of 57. BMI, TyG, and TyG-BMI indices increased significantly (23.16 to 23.51 for BMI; 8.59 to 8.62 for TyG, 199.14 to 204.22 for TyG-BMI) over 4 years. Participants were categorized based on cumulative TyG and TyG-BMI medians at 25.84 and 606.43, respectively. In the lower TyG group, each SBP unit increment raised CVD risk by 0.5% at 4 years (HR = 1.005, 95% CI 1.003–1.007, pConclusion: Cumulative SBP and DBP increases are linked to higher CVD risks, especially in individuals with lower cumulative TyG or TyG-BMI indices. DBP effects were more substantial, particularly noticeable over 9 years. This study underscores the nuanced impacts of blood pressure and metabolic indicators on long-term cardiovascular health.
- 1 September 2024
- Published by Wolters Kluwer Health in Hypertension
- Vol. 81 (Suppl_1) , AP135
Abstract P135: Evaluating the Impact of Cumulative Blood Pressure and Metabolic Indicators on Long-Term Cardiovascular Risk: Insights from a 9-year Nationwide Prospective Cohort Study
Background: Cardiovascular diseases (CVD) are a global health challenge, particularly for older adults. Blood pressure and the Triglyceride-Glucose (TyG) index, an insulin resistance indicator, are key predictors of cardiometabolic risks. Methods: Using data from the China Health and Retirement Longitudinal Study (CHARLS), this study followed participants without baseline CVD over 9 years. We stratified subjects by cumulative TyG and TyG-BMI indices, applying Cox proportional hazards adjusted by age and gender, and restricted cubic spline regression for CVD risk evaluation. Results: From 11,847 participants, 4,751 were analyzed, mainly females (55.25%), with a median age of 57. BMI, TyG, and TyG-BMI indices increased significantly (23.16 to 23.51 for BMI; 8.59 to 8.62 for TyG, 199.14 to 204.22 for TyG-BMI) over 4 years. Participants were categorized based on cumulative TyG and TyG-BMI medians at 25.84 and 606.43, respectively. In the lower TyG group, each SBP unit increment raised CVD risk by 0.5% at 4 years (HR = 1.005, 95% CI 1.003–1.007, pConclusion: Cumulative SBP and DBP increases are linked to higher CVD risks, especially in individuals with lower cumulative TyG or TyG-BMI indices. DBP effects were more substantial, particularly noticeable over 9 years. This study underscores the nuanced impacts of blood pressure and metabolic indicators on long-term cardiovascular health.
- Open Access
- 20 June 2024
- Published by Frontiers Media SA in Frontiers in Cellular and Infection Microbiology
- Vol. 14, 1429197
Editorial: Machine learning and deep learning applications in pathogenic microbiome research
- 20 June 2024
- Published by Frontiers Media SA in Frontiers in Cellular and Infection Microbiology
- Vol. 14, 1429197
Open AccessEditorial: Machine learning and deep learning applications in pathogenic microbiome research
- Top Cited Papers
Global Burden of Cardiovascular Diseases and Risks, 1990-2022
- Open Access
Gut microbiota in heart failure and related interventions
Heart failure (HF) is a sophisticated syndrome with structural or functional impairment of ventricular filling or ejection of blood, either causing symptoms and signs or being asymptomatic. HF is a major global health issue affecting about 64.3 million people worldwide. The gut microbiota refers to the complex ecosystem of microorganisms, mainly bacteria, in the gut. Studies have revealed that the gut microbiota is associated with many diseases ranging from neurodegenerative diseases to inflammatory bowel disease and cardiovascular diseases. The gut hypothesis of HF suggests that low cardiac output and systemic circulation congestion would cause insufficient intestinal perfusion, leading to ischemia and intestinal barrier dysfunction. The resulting bacterial translocation would contribute to inflammation. Recent studies have refined the hypothesis that changes of metabolites in the gut microbiota have a close relationship with HF. Thus, the gut microbiota has emerged as a potential therapeutic target for HF due to both its critical role in regulating host physiology and metabolism and its pivotal role in the development of HF. This review article aims to provide an overview of the current understanding of the gut microbiota's involvement in HF, including the introduction of the gut hypothesis of HF, its association with HF progression, the potential mechanisms involved mediated by the gut microbiota metabolites, and the impact of various interventions on the gut microbiota, including dietary interventions, probiotic therapy, fecal microbiota transplantation, antibiotics, and so on. While the gut hypothesis of HF is refined with up‐to‐date knowledge and the gut microbiota presents a promising target for HF therapy, further research is still needed to further understand the underlying mechanisms between gut microbiota and HF, the efficacy of these interventions, and contribute to the health of HF patients.
Open AccessGut microbiota in heart failure and related interventions
Heart failure (HF) is a sophisticated syndrome with structural or functional impairment of ventricular filling or ejection of blood, either causing symptoms and signs or being asymptomatic. HF is a major global health issue affecting about 64.3 million people worldwide. The gut microbiota refers to the complex ecosystem of microorganisms, mainly bacteria, in the gut. Studies have revealed that the gut microbiota is associated with many diseases ranging from neurodegenerative diseases to inflammatory bowel disease and cardiovascular diseases. The gut hypothesis of HF suggests that low cardiac output and systemic circulation congestion would cause insufficient intestinal perfusion, leading to ischemia and intestinal barrier dysfunction. The resulting bacterial translocation would contribute to inflammation. Recent studies have refined the hypothesis that changes of metabolites in the gut microbiota have a close relationship with HF. Thus, the gut microbiota has emerged as a potential therapeutic target for HF due to both its critical role in regulating host physiology and metabolism and its pivotal role in the development of HF. This review article aims to provide an overview of the current understanding of the gut microbiota's involvement in HF, including the introduction of the gut hypothesis of HF, its association with HF progression, the potential mechanisms involved mediated by the gut microbiota metabolites, and the impact of various interventions on the gut microbiota, including dietary interventions, probiotic therapy, fecal microbiota transplantation, antibiotics, and so on. While the gut hypothesis of HF is refined with up‐to‐date knowledge and the gut microbiota presents a promising target for HF therapy, further research is still needed to further understand the underlying mechanisms between gut microbiota and HF, the efficacy of these interventions, and contribute to the health of HF patients.