Machine Learning for the ECG Diagnosis and Risk Stratification of Occlusion Myocardial Infarction at First Medical...

Area of Science:

• Cardiology
• Artificial Intelligence
• Medical Diagnostics

Background:

• Occlusion myocardial infarction (OMI) without ST-elevation presents a diagnostic challenge, impacting patient prognosis.
• Current triage tools lack accuracy in identifying these high-risk patients for immediate reperfusion therapy.

Approach:

• Developed and validated machine learning models using a large, multi-site observational cohort of 7,313 patients.
• The study focused on ECG diagnosis of OMI, aiming to improve precision and sensitivity.

Key Points:

• The derived intelligent model significantly outperformed clinicians and commercial systems in OMI diagnosis.
• An OMI risk score demonstrated superior rule-in and rule-out accuracy compared to standard care.
• Integrating the OMI risk score with clinical judgment reclassified one-third of chest pain patients.

Conclusions:

• Machine learning offers a powerful tool for the ECG diagnosis of OMI in non-ST-elevation patients.
• The validated OMI risk score can enhance clinical decision-making and patient management.
• ECG features identified by the model provide mechanistic insights into myocardial injury.