AID-OMIE - Artificial Intelligence in Detection of Occlusive Myocardial Infarction in Emergency Medicine

Study Objective and Hypothesis The study hypothesizes that artificial intelligence (AI)-assisted interpretation of the 12-lead electrocardiogram (ECG) can improve the care of patients resuscitated after out-of-hospital cardiac arrest (OHCA) by enabling faster and more accurate detection of occlusion myocardial infarction (OMI). This enhanced diagnostic approach could reduce the time required for revascularization, improve patient outcomes, and decrease unnecessary activations of cardiac catheterization laboratories. The primary objective of the study is to assess the effectiveness of an AI-powered ECG model in identifying acute OMI in OHCA patients whose post-return of spontaneous circulation (ROSC) ECG does not show ST-elevation. Methods This is a retrospective observational study involving OHCA patients in Bolzano, Italy, who meet the following inclusion criteria: Aged 18 years or older. Achieved ROSC after cardiac arrest. Underwent coronary angiography (CAG) within seven days post-OHCA. Prehospital post-ROSC ECG and CAG reports available. Exclusion criteria include in-hospital cardiac arrest (IHCA), traumatic cardiac arrest, cardiac arrest from a non-cardiac cause, and poor-quality or corrupted ECG images. Post-ROSC ECGs will be analyzed using the PMcardio App, an AI tool for ECG interpretation. The data will be fully anonymized before storage. Coronary angiography charts will be reviewed for the presence of atherosclerotic lesions, the degree of arterial narrowing, and Thrombolysis in Myocardial Infarction (TIMI) flow, which assesses blood flow in coronary arteries. Study Outcomes The primary outcome is the sensitivity and specificity of the AI-assisted ECG in detecting OMI in patients whose post-ROSC ECG does not show ST-elevation. Secondary outcomes include the frequency of OMI in OHCA patients without ST-elevation and the ability of the AI model to rule out OMI accurately in these cases.

Study Objective and Hypothesis The study hypothesizes that artificial intelligence (AI)-assisted interpretation of the 12-lead electrocardiogram (ECG) can improve the care of patients resuscitated after out-of-hospital cardiac arrest (OHCA) by enabling faster and more accurate detection of occlusion myocardial infarction (OMI). This enhanced diagnostic approach could reduce the time required for revascularization, improve patient outcomes, and decrease unnecessary activations of cardiac catheterization laboratories. The primary objective of the study is to assess the effectiveness of an AI-powered ECG model in identifying acute OMI in OHCA patients whose post-return of spontaneous circulation (ROSC) ECG does not show ST-elevation. Methods This is a retrospective observational study involving OHCA patients in Bolzano, Italy, who meet the following inclusion criteria: OHCA from 2018-2025 Aged 18 years or older. Achieved ROSC after cardiac arrest. Underwent coronary angiography (CAG) within seven days post-OHCA. Prehospital post-ROSC ECG and CAG reports available. Exclusion criteria include in-hospital cardiac arrest (IHCA), traumatic cardiac arrest, cardiac arrest from a non-cardiac cause, and poor-quality or corrupted ECG images. Post-ROSC ECGs will be analyzed using the PMcardio App, an AI tool for ECG interpretation. The data will be fully anonymized before storage. Coronary angiography charts will be reviewed for the presence of atherosclerotic lesions, the degree of arterial narrowing, and Thrombolysis in Myocardial Infarction (TIMI) flow, which assesses blood flow in coronary arteries. Study Outcomes The primary outcome is the sensitivity and specificity of the AI-assisted ECG in detecting OMI in patients whose post-ROSC ECG does not show ST-elevation. Secondary outcomes include the frequency of OMI in OHCA patients without ST-elevation and the ability of the AI model to rule out OMI accurately in these cases.