Diagnostic Model of In-Hospital Mortality in Patients with Acute ST-Segment Elevation Myocardial Infarction Used...

Area of Science:

• Cardiology
• Medical Informatics
• Artificial Intelligence

Background:

• Preventing in-hospital mortality in ST-segment elevation myocardial infarction (STEMI) is critical.
• Acute STEMI poses significant risks to patient survival during hospitalization.

Purpose of the Study:

• To develop and externally validate a diagnostic model for in-hospital mortality in acute STEMI patients.
• To leverage artificial intelligence methods for improved mortality prediction.

Main Methods:

• Utilized AI methods to develop and validate diagnostic models using training, testing, and validation datasets.
• Employed confusion matrix and Area Under the Receiver Operating Characteristic Curve (AUC) for model evaluation.
• Balanced unbalanced data to enhance model robustness.

Main Results:

• Identified key predictors of mortality: age, gender, cardiogenic shock, atrial fibrillation (AF), ventricular fibrillation (VF), third-degree atrioventricular block, bleeding, PCI, CABG, hypertension, diabetes, and prior MI.
• Logistic regression model achieved an F2 score of 0.81 (training), 0.6 (test), and 0.59 (validation).
• Logistic regression model achieved an AUC of 0.77 (training), 0.78 (test), and 0.8 (validation), indicating strong performance.

Conclusions:

• The diagnostic model built using logistic regression proved to be the most effective.
• The study successfully developed and validated AI-driven models for predicting in-hospital mortality in STEMI patients.
• This research provides a valuable tool for risk stratification and management of STEMI patients.