Machine learning-based prediction model for myocardial ischemia under high altitude exposure: a cohort study.

Area of Science:

• Cardiology
• Aerospace Medicine
• Data Science

Background:

• High altitude exposure is a known risk factor for myocardial ischemia (MI) and cardiovascular mortality.
• Existing cardiovascular disease prediction models lack specificity for high-altitude induced MI.
• There is a need for predictive tools to identify individuals at risk of MI in high-altitude environments.

Purpose of the Study:

• To develop and validate a machine learning-based prediction model for high-altitude myocardial ischemia (MI).
• To identify key laboratory risk factors associated with high-altitude MI.
• To provide a tool for early risk assessment before high-altitude exposure.

Main Methods:

• A prospective cohort study involving soldiers undergoing high-altitude training.
• Consolidation of health examination and questionnaire data from two distinct periods.
• Development of a predictive model using a training dataset (n=2141) and validation on a test dataset (n=714) via random assignment (3:1 ratio).
• Model performance evaluated using the area under the receiver operating characteristic curve (AUC).

Main Results:

• A highly accurate machine learning model for predicting high-altitude MI was developed (AUC=0.86).
• The model identified five key laboratory predictors: Eosinophils percentage (Eos.Per), Globulin (G), Calcium (Ca), Glucose (GLU), and Aspartate aminotransferase (AST).
• The model demonstrated strong predictive performance on the independent test dataset.

Conclusions:

• A concise and accurate machine learning model based on five laboratory results can predict high-altitude MI incidence.
• The model can be utilized for proactive risk identification, enabling better preparation for high-altitude environments.
• Further external validation across diverse demographics, including females and various age groups, is recommended.