Presentation Schedule
Machine Learning Approach to Develop a Risk Prediction Model Based on Minor Physical Anomalies for Psychological Resilience in Schizophrenia (89058)
Wednesday, 26 March 2025 15:40
Session: Poster Session 3
Room: Orion Hall (5F)
Presentation Type: Poster Presentation
Psychological resilience is the ability to adapt emotionally and socially to stress, adversity, or trauma. Minor physical anomalies (MPA) are subtle abnormal features of the head, eyes, ears, mouth, hands, and feet. Studies have shown that more MPA is linked to more severe schizophrenia symptoms, and lower resilience is associated with worse outcomes in schizophrenia. This study aimed to use machine learning algorithms to identify MPA variables distinguishing between low and high resilience in schizophrenia patients. We enrolled 163 schizophrenia patients admitted to hospitals and used the Connor-Davidson Resilience Scale to categorize them into low and high resilience groups. We utilized variable selection using random forests (varSelRF) to identify the important MPA variables. Logistic regression and three machine learning algorithms (random forest, support vector machine, and eXtreme gradient boosting) were used to develop the risk prediction model. The 14 important MPA variables were selected by varSelRF. The results showed that using these important MPAs variables, along with sex and BMI, to develop the risk prediction model for discriminating low and high resilience schizophrenia resulted in an Area Under the Curve (AUC) of 0.70-0.81. We then used a stacked ensemble model to combine the predictions of four models, achieving an AUC value of 0.83 (sensitivity = 0.86, specificity = 0.65). The MPAs can serve as neurodevelopment markers to predict resilience in schizophrenia. The risk prediction models provide a clinical decision support system for detecting low or high resilience in schizophrenia patients, enabling early intervention in clinical practice.
Authors:
Sheng-Hsiang Lin, National Cheng Kung University, Taiwan
Chih-Wei Lin, National Cheng Kung University, Taiwan
Jin-Jia Lin, Chi Mei Medical Center, Taiwan
Huai-Hsuan Tseng, National Cheng Kung University Hospital, Taiwan
Chih-Chun Huang, National Cheng Kung University Hospital, Taiwan
Shulan Hsieh, National Cheng Kung University, Taiwan
About the Presenter(s)
Dr Sheng-Hsiang Lin is currently at National Cheng Kung University in Taiwan
See this presentation on the full schedule – Wednesday Schedule
Conference Comments & Feedback
Place a comment using your LinkedIn profileShare this Presentation
Comments
Powered by WP LinkPress