A Survival Prediction Model of Self-Immolation Based on Machine Learning Techniques.

Area of Science:

• Medical Informatics
• Computational Medicine
• Public Health

Background:

• Self-immolation is a violent suicide method prevalent in developing nations.
• Accurate survival prediction is crucial for effective therapeutic strategies in self-immolation cases.
• Machine learning (ML) offers advanced tools for disease diagnosis and patient survival prediction.

Purpose of the Study:

• To develop and evaluate a machine learning model for predicting the survival of self-immolation patients.
• To identify key factors influencing patient survival.
• To provide a data-driven tool for clinical decision-making in burn centers.

Main Methods:

• A retrospective cross-sectional study analyzed 445 self-immolated patients (2008-2019).
• Multiple ML algorithms, including Gradient Boosting, SVM, random forest, MLP, and KNN, were implemented using Python 3.7.
• Model performance was assessed using F1 score, accuracy, sensitivity, specificity, and AUC.

Main Results:

• The Support Vector Machine (SVM) model demonstrated superior performance with an F1 score of 91.8%, accuracy of 91.9%, and AUC of 0.96.
• Key predictors of survival identified by the ML model include surgical procedures, patient score, length of stay, anatomical region, and gender.
• These factors significantly impacted survival prediction more than others.

Conclusions:

• Machine learning algorithms can effectively predict self-immolation patient survival.
• The developed SVM model provides a valuable tool for predicting outcomes and guiding treatment.
• Findings can inform clinical management, policy development, and decision-making in burn care settings.