Development of a Machine Learning Model for Diagnosing Occlusive Myocardial Infarction in Patients with Left Bundle Branch Block

What this Trial Is About

Researchers are evaluating a new method to diagnose severe heart attacks in patients who have a specific heart electrical pattern called Left Bundle Branch Block (LBBB). Diagnosing heart attacks in these patients is challenging because LBBB changes the heart's electrical signals on an ECG, hiding typical signs of blocked arteries. This observational study aims to develop and test a machine learning model that analyzes digital 12-lead ECG signals to detect true artery blockages in patients with LBBB. The study uses standard digital 12-lead ECG data collected during emergency visits and applies a machine learning model to analyze these signals for subtle patterns. The model's predictions will be compared with results from invasive coronary angiography, the gold standard for visualizing blocked vessels. The study also evaluates whether the model can distinguish between heart attacks caused by blocked arteries (Type 1 MI) and other conditions with elevated heart enzymes (Type 2 MI). Participants will have their ECG recorded when they arrive at the emergency department with chest pain or related symptoms. Researchers will compare the machine learning model's results with coronary angiography performed during the hospital stay. The main outcome measures include the accuracy of diagnosing occlusive heart attacks within about 24 hours and the ability to differentiate heart attack types during the hospital stay. Researchers will also assess whether using this tool can reduce unnecessary invasive procedures. Participation involves data collection during routine care and follow-up during the hospital stay, with the study ending after these assessments.

Machine Learning for Diagnosis of Occlusive MI in LBBB Patients