GHGuoping Huang

Guoping Huang

*This is an algorithmically generated author record based on publicly available information.

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


Guoping Huang has a total of 67 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    The complexity of associations between borderline personality features, depression symptoms, anxiety symptoms, and suicidal behavior in adolescents: a study based on network analysis

    by Chunyou Li, Shuhan He, Jian Wen, Shibei Zhang, Linhong Chen, Yaxin Hu, Lingjiang Liu, Jie Gao, Junlin Wu, Guoping Huang

  • Open Access

    Association between sleep disorders and diabetic kidney disease among U.S. adults aged 45 and older: findings from the 2007–2014 NHANES

    by Lingjiang Liu, Shibei Zhang, Junlin Wu, Jie Gao, Chunyou Li, Shuhan He, Yaxin Hu, Huaiyi Zhu, Guoping Huang

    Background: Diabetic kidney disease (DKD) is a frequent microvascular manifestation of diabetes. However, the associations between sleep disorders, sleep duration time (SDT), and DKD remain inadequately understood. This study seeks to examine the associations of sleep-related factors with the prevalence of DKD among Americans aged 45 and older. Method: A cross-sectional analysis was carried out utilizing the 2007–2014 National Health and Nutrition Examination Survey (NHANES) data. Four sleep-related variables were evaluated: continuous SDT, categorized SDT, self-reported sleep disorders, and clinically diagnosed sleep disorders. The associations of sleep-related factors with DKD were examined via weighted logistic regression models and subgroup analyses, with adjustments for multiple potential confounders. Possible nonlinear associations between SDT and DKD were explored using restricted cubic spline (RCS) functions. Propensity score matching (PSM) was applied to evaluate the robustness of the findings. Results: 10,780 participants were included. The overall DKD prevalence was 11.5% (n = 1243). Both self-reported sleep disorders (Model 3: OR = 1.43, 95% CI 1.16–1.77, P < 0.05) and clinically diagnosed sleep disorders (Model 3: odds ratio (OR) = 1.56, 95% confidence interval (CI) 1.15–2.11, P < 0.05) were associated with a higher prevalence of DKD. Prolonged SDT (> 9 h) was also associated with a higher DKD prevalence (OR = 1.63, 95% CI 1.08–2.46, P < 0.05). RCS analysis demonstrated a U-shaped association between SDT and DKD (P for overall association = 0.001; P for nonlinearity < 0.001). Conclusion: In this cross-sectional analysis, sleep disorders and abnormal sleep duration were associated with DKD. Sleep-related factors may warrant consideration in the comprehensive clinical assessment of individuals with diabetes. Longitudinal studies are required to clarify temporal relationships.

  • Open Access

    Large Language Models for Psychiatric Diagnosis Based on Multicenter Real-World Clinical Records: Comparative Study

    by Maoqian Sun, Jia Yu, Zhuhong Long, Yun Yang, Tao Xiao, Jiaquan Liang, Jun Feng, Huaili Deng, Guoping Huang

    Background: Psychiatric disorders are diagnostically challenging and often rely on subjective clinical judgment, particularly in resource-limited settings. Large language models (LLMs) have demonstrated potential in supporting psychiatric diagnosis; however, robust evidence from large-scale, real-world clinical data remains limited. Objective: This study aimed to evaluate and compare the diagnostic performance of multiple LLMs for psychiatric disorders using multicenter real-world electronic health records (EHRs). Methods: We retrospectively analyzed 9923 inpatient EHRs collected from 6 psychiatric centers across China, encompassing all ICD-10 (International Statistical Classification of Diseases, Tenth Revision) psychiatric categories. In total, 3 LLMs—GPT-4.0 (OpenAI), GPT-3.5 (OpenAI), and GLM-4-Plus (Zhipu AI)—were evaluated against physician-confirmed discharge diagnoses. Diagnostic performance was assessed using strict accuracy criteria and lenient classification metrics, with subgroup analyses conducted across diagnostic categories and age groups. Results: GPT-4.0 achieved the highest overall strict diagnostic accuracy (71.7%) and the highest weighted F1-score under lenient evaluation (0.881), particularly for high-prevalence disorders, such as mood disorders and schizophrenia spectrum disorders. Diagnostic performance varied across age groups, with the highest accuracy observed in older adult patients (up to 79.5%) and lower accuracy in adolescents. Across centers, model performance remained stable, with no significant intercenter differences. Conclusions: LLMs—especially GPT-4.0—demonstrate promising capability in supporting psychiatric diagnosis using real-world EHRs. However, diagnostic performance varies by age group and disorder category. LLMs should be regarded as assistive tools rather than replacements for clinical judgment, and further validation is needed before routine clinical implementation.


  • Open Access

    The complexity of associations between emotion regulation, interpersonal sensitivity, cognitive insight, and non-suicidal self-injury: a study based on network analysis

    by Shuhan He, Chunyou Li, Shibei Zhang, Yaxin Hu, Jian Wen, Lingjiang Liu, Jie Gao, Junlin Wu, Guoping Huang

    Adolescence is a period marked by emotional volatility, interpersonal vulnerability, and underdeveloped cognitive control, making youths especially susceptible to non-suicidal self-injury. Distinct forms of NSSI may differ in psychological function and clinical relevance. This study is the first to employ network analysis with dual network models to examine the complex interplay among emotion regulation, interpersonal sensitivity, cognitive insight, and 18 specific NSSI behaviors in adolescents. The primary goal was to identify key psychological nodes and central NSSI behaviors within the network structure to inform more targeted prevention and intervention strategies. A total of 5572 adolescents (ages 12–18; 53.9% male) from three urban secondary schools in Nanning, Guangxi Zhuang Autonomous Region, China, participated in self-report surveys conducted between September and November 2023. Standardized and previously validated instruments were used to assess NSSI behaviors, emotion regulation, interpersonal sensitivity, and cognitive insight. Network analysis using mixed graphical models and LASSO regularization was conducted to identify central and bridging nodes across binary-level and symptom-level NSSI networks. Network analysis revealed robust interconnections between emotion regulation, interpersonal sensitivity, cognitive insight, and various NSSI behaviors. In the binary-level network, the fragile inner-self (strength = 1.5) dimension of interpersonal sensitivity exhibited the highest strength, while cognitive reappraisal(bridge strength = 0.88) and expressive suppression(bridge strength = 0.73) showed the strongest bridge strength. Self-reflection was directly associated with NSSI behaviors, and self-certainty was linked to both emotion regulation and interpersonal sensitivity. In the symptom-level network, the most central NSSI behaviors included deliberately scraping the skin to cause bleeding, tying objects around the body, punching hard surfaces, banging the head, and cutting the skin. Fragile inner-self (bridge strength = 0.79) emerged as the strongest bridge node in this network. This study suggests that emotion regulation, interpersonal sensitivity and cognitive insight are closely associated with adolescent NSSI behaviors, highlighting their potential relevance for intervention strategies.


  • Open Access

    Discovery of factors associated with smartphone addiction among high school adolescents: Using machine learning and network analysis

    by Zhiyuan Li, Qianmei Long, Menghan Hao, Lingjiang Liu, Yujia Weng, Congwen Yang, Junlin Wu, Bin Huang, Junlin Qiu, Yanqing Huang, et al.


  • Top Cited Papers
    Open Access

    Predictive Performance of Machine Learning for Suicide in Adolescents: Systematic Review and Meta-Analysis

    by Lingjiang Liu, Zhiyuan Li, Yaxin Hu, Chunyou Li, Shuhan He, Shibei Zhang, Jie Gao, Huaiyi Zhu, Guoping Huang

    Background: In the context of escalating global mental health challenges, adolescent suicide has become a critical public health concern. In current clinical practices, considerable challenges are encountered in the early identification of suicide risk, as traditional assessment tools demonstrate limited predictive accuracy. Recent advancements in machine learning (ML) present promising solutions for risk prediction. However, comprehensive evaluations of their efficacy in adolescent populations remain insufficient. Objective: This study systematically assessed the performance of ML-based prediction models across various suicide-related behaviors in adolescents, aiming to establish an evidence-based foundation for the development of clinically applicable risk assessment tools. Methods: This review assessed ML for predicting adolescent suicide–related behaviors. PubMed, Embase, Cochrane, and Web of Science databases were rigorously searched until April 20, 2024, and a multivariate prediction model was employed to assess the risk of bias. The c-index was used as the primary outcome measure to conduct a meta-analysis on nonsuicidal self-injury (NSSI), suicidal ideation, suicide attempts, suicide attempts combined with suicidal ideation, and suicide attempts combined with NSSI, evaluating their accuracy in the validation set. Results: A total of 42 studies published from 2018 to 2024 were included, encompassing 104 distinct ML models and 1,408,375 adolescents aged 11 to 20 years. The combined area under the receiver operating characteristic curve values for ML models in predicting NSSI, suicidal ideation, suicide attempts, suicide attempts combined with suicidal ideation, and suicide attempts combined with NSSI were 0.79 (95% CI 0.72-0.86), 0.77 (95% CI 0.71-0.83), 0.84 (95% CI 0.83-0.86), 0.82 (95% CI 0.79-0.84), and 0.75 (95% CI 0.73-0.76), respectively. The ML models demonstrated the highest combined sensitivity for suicide attempt prediction, with a value of 0.80 (95% CI 0.75-0.84), and the highest combined specificity for NSSI prediction, with a value of 0.96 (95% CI 0.94-0.99). Conclusions: Our findings suggest that ML techniques exhibit promising predictive performance for forecasting suicide risk in adolescents, particularly in predicting suicide attempts. Notably, ensemble methods, such as random forest and extreme gradient boosting, showed superior performance across multiple outcome types. However, this study has several limitations, including the predominance of internal validation methods employed in the included literature, with few studies employing external validation, which may limit the generalizability of the results. Future research should incorporate larger and more diverse datasets and conduct external validation to improve the prediction capability of these models, ultimately contributing to the development of ML-based adolescent suicide risk prediction tools.


  • Open Access

    Large Language Models for Psychiatric Diagnosis Based on Multicenter Real-World Clinical Records: Comparative Study (Preprint)

    by Maoqian Sun, Jia Yu, Zhuhong Long, Yun Yang, Tao Xiao, Jiaquan Liang, Jun Feng, Huaili Deng, Guoping Huang

    BACKGROUND Psychiatric disorders are diagnostically challenging and often rely on subjective clinical judgment, particularly in resource-limited settings. Large language models (LLMs) have demonstrated potential in supporting psychiatric diagnosis; however, robust evidence from large-scale, real-world clinical data remains limited. OBJECTIVE This study aimed to evaluate and compare the diagnostic performance of multiple LLMs for psychiatric disorders using multicenter real-world electronic health records (EHRs). METHODS We retrospectively analyzed 9923 inpatient EHRs collected from 6 psychiatric centers across China, encompassing all ICD-10 (International Statistical Classification of Diseases, Tenth Revision) psychiatric categories. In total, 3 LLMs—GPT-4.0 (OpenAI), GPT-3.5 (OpenAI), and GLM-4-Plus (Zhipu AI)—were evaluated against physician-confirmed discharge diagnoses. Diagnostic performance was assessed using strict accuracy criteria and lenient classification metrics, with subgroup analyses conducted across diagnostic categories and age groups. RESULTS GPT-4.0 achieved the highest overall strict diagnostic accuracy (71.7%) and the highest weighted F1-score under lenient evaluation (0.881), particularly for high-prevalence disorders, such as mood disorders and schizophrenia spectrum disorders. Diagnostic performance varied across age groups, with the highest accuracy observed in older adult patients (up to 79.5%) and lower accuracy in adolescents. Across centers, model performance remained stable, with no significant intercenter differences. CONCLUSIONS LLMs—especially GPT-4.0—demonstrate promising capability in supporting psychiatric diagnosis using real-world EHRs. However, diagnostic performance varies by age group and disorder category. LLMs should be regarded as assistive tools rather than replacements for clinical judgment, and further validation is needed before routine clinical implementation.

  • Open Access

    Predictive Performance of Machine Learning for Suicide in Adolescents: Systematic Review and Meta-Analysis (Preprint)

    by Lingjiang Liu, Zhiyuan Li, Yaxin Hu, Chunyou Li, Shuhan He, Shibei Zhang, Jie Gao, Huaiyi Zhu, Guoping Huang

    BACKGROUND In the context of escalating global mental health challenges, adolescent suicide has become a critical public health concern. In current clinical practices, considerable challenges are encountered in the early identification of suicide risk, as traditional assessment tools demonstrate limited predictive accuracy. Recent advancements in machine learning (ML) present promising solutions for risk prediction. However, comprehensive evaluations of their efficacy in adolescent populations remain insufficient. OBJECTIVE This study systematically assessed the performance of ML-based prediction models across various suicide-related behaviors in adolescents, aiming to establish an evidence-based foundation for the development of clinically applicable risk assessment tools. METHODS This review assessed ML for predicting adolescent suicide–related behaviors. PubMed, Embase, Cochrane, and Web of Science databases were rigorously searched until April 20, 2024, and a multivariate prediction model was employed to assess the risk of bias. The c-index was used as the primary outcome measure to conduct a meta-analysis on nonsuicidal self-injury (NSSI), suicidal ideation, suicide attempts, suicide attempts combined with suicidal ideation, and suicide attempts combined with NSSI, evaluating their accuracy in the validation set. RESULTS A total of 42 studies published from 2018 to 2024 were included, encompassing 104 distinct ML models and 1,408,375 adolescents aged 11 to 20 years. The combined area under the receiver operating characteristic curve values for ML models in predicting NSSI, suicidal ideation, suicide attempts, suicide attempts combined with suicidal ideation, and suicide attempts combined with NSSI were 0.79 (95% CI 0.72-0.86), 0.77 (95% CI 0.71-0.83), 0.84 (95% CI 0.83-0.86), 0.82 (95% CI 0.79-0.84), and 0.75 (95% CI 0.73-0.76), respectively. The ML models demonstrated the highest combined sensitivity for suicide attempt prediction, with a value of 0.80 (95% CI 0.75-0.84), and the highest combined specificity for NSSI prediction, with a value of 0.96 (95% CI 0.94-0.99). CONCLUSIONS Our findings suggest that ML techniques exhibit promising predictive performance for forecasting suicide risk in adolescents, particularly in predicting suicide attempts. Notably, ensemble methods, such as random forest and extreme gradient boosting, showed superior performance across multiple outcome types. However, this study has several limitations, including the predominance of internal validation methods employed in the included literature, with few studies employing external validation, which may limit the generalizability of the results. Future research should incorporate larger and more diverse datasets and conduct external validation to improve the prediction capability of these models, ultimately contributing to the development of ML-based adolescent suicide risk prediction tools.

  • Open Access

    Peer victimization and non-suicidal self-injury among high school students: the mediating role of social anxiety, mobile phone addiction, and sex differences

    by Qianmei Long, Bin Huang, Yiyu Tang, Junlin Wu, Jia Yu, Junlin Qiu, Yanqing Huang, Guoping Huang

    Background: Peer victimization (PV) is one of the major causes of non-suicidal self-injury. Non-suicidal self-injury (NSSI), peer victimization, social anxiety, and mobile phone addiction are significantly related; however, the interaction mechanism and effect of sex differences remain to be determined. Objective: Herein, we investigated the relationship between peer victimization and NSSI among Chinese high school students. We also explored the chain mediating roles of social anxiety and mobile phone addiction and the regulatory role of sex. The findings of this study provide insights for theoretical interventions based on internal mechanisms. Method: A self-reported survey of 14,666 high school students from Sichuan County was conducted using a peer victimization scale, NSSI scale, social anxiety scale, and mobile phone addiction scale. A self-administered questionnaire was used to capture sociodemographic information. Results: Peer victimization, social anxiety, and mobile phone addiction were positively correlated with NSSI. Peer victimization had significant direct predictive effects on NSSI (95% CI: 0.341, 0.385) and significant indirect predictive effects on NSSI through social anxiety (95% CI: 0.008, 0.019) or mobile phone addiction (95% CI: 0.036, 0.053). Peer victimization had significant indirect predictive effects on NSSI through social anxiety as well as mobile phone addiction (95% CI: 0.009, 0.014). The first stage (predicting the effect of peer victimization on NSSI) and the third stage (predicting the effect of mobile phone addiction on NSSI) were both moderated by sex. Conclusions: Peer victimization could directly predict NSSI and indirectly predict NSSI through social anxiety and mobile phone addiction. Thus, social anxiety and mobile phone addiction exhibited chain mediating effects between peer victimization and NSSI in high school students; moreover, sex might be involved in the regulation of the mediation process.


  • Stressful life events associate with adolescent depression through dysfunction of frontoparietal network

    by Yingxue Gao, Ruohan Feng, Lihua Zhuo, Kaili Liang, Weijie Bao, Hui Qiu, Zilin Zhou, Guoping Huang, Xiaoqi Huang

    The current study investigate functional network alterations in adolescent MDD using the clustering analysis. We found that adolescents with MDD showed decreased connectivity between the frontoparietal network (FPN) and other networks relative to HC. The global connectivity strength of FPN was negatively correlated with the punishment and academic pressure factors, and mediated the association between these two stressors and depression. Our findings suggest that punishment and academic pressure may contribute to the development of depression in adolescence through dysfunction of the FPN, which may promote future prevention of adolescent MDD.