AI-Powered ECG Detecting Culprit Vessel Blood Flow Abnormality in ACS

Recruiting

• Conditions: STEMI; Coronary Syndrome; Coronary Occlusion; NSTEMI; NSTEMI - Non-ST Segment Elevation MI; Acute Coronary Syndrome

• Registration Number: NCT06528821
• Lead Sponsor: Powerful Medical

Brief Summary

AI ECG TIMI is an investigator-initiated, international, and multicenter registry of acute coronary syndrome patients aimed to identify electrocardiographic findings detected by an AI model predicting coronary blood flow alteration. The aim of the study is to identify electrocardiographic findings detected by an automated artificial intelligence (AI) model that can predict coronary blood flow alteration as assessed by the TIMI grade flow at the very moment of the invasive coronary angiography in patients with acute coronary syndromes.

Detailed Description

Not available

Study Timeline

• Study First Submitted: 2024-07-26
• Study First Submitted QC: 2024-07-26
• Study First Posted: 2024-07-30
• Study Start: 2024-12-10
• Status Verified: 2025-04
• Last Update Submitted: 2025-04-24
• Last Update Posted: 2025-04-29
• Primary Completion: 2025-10-01
• Study Completion: 2025-10-01

Recruitment & Eligibility

• Status: RECRUITING
• Sex: All
• Target Recruitment: 800

Inclusion Criteria

• Patients with acute coronary syndromes undergoing invasive coronary angiography as clinically indicated.
• Availability of a standard 12-lead ECG performed at invasive coronary angiogram before percutaneous coronary intervention (maximum within 15 min before vascular access).
• Age ≥18 years.

Exclusion Criteria

• Individuals presenting for a non-emergent (elective) cardiac catheterization.
• Individuals presenting with chronic coronary syndrome (CCS) or stable angina symptoms.
• Individuals without symptoms suspicious for acute coronary syndromes.
• Individuals with contraindications for cardiac catheterization.

Study & Design

• Study Type: OBSERVATIONAL
• Study Design: Not specified

Primary Outcome Measures

Name	Time	Method
Diagnostic Accuracy of the AI-ECG TIMI AI Model	Index hospitalization (assessed up to 5 days)	The primary endpoint is the AI-ECG TIMI model's ability to identify patients with actively occluded (TIMI 0-1) at the time of invasive coronary angiography using only single-standard 12-lead ECGs assessed by accuracy, sensitivity, specificity, positive, negative predictive values and F1 scores.

Secondary Outcome Measures

Name	Time	Method
STEMI-equivalent ECG patterns	Index hospitalization (assessed up to 5 days)	a. Prevalence of STEMI equivalent ECG patterns and their correlation to TIMI flow grade at angiography.
Subgroup analysis of NSTEMI-OMI patients	Index hospitalization (assessed up to 5 days)	9. NSTEMI-OMI stratified according to immediate (\<2h) vs. SoC (\>2h) invasive management, median: i. Peak troponin (maximum rise before fall) ii. Median LVEF at discharge iii. Length of hospital stay iv. Length of cardiac care unit (CCU) or intensive care unit (ICU) stay v. Presence of life-threatening arrhythmias
Subgroup Performance	Index hospitalization (assessed up to 5 days)	AI-ECG TIMI AI Model's performance across subgroups; 1. Demographic; 2. Risk factors \& co-morbidities; 3. Electrocardiographic; 4. Culprit artery territory
Correlate established definitions of occlusion myocardial infarction (OMI) to myocardial perfusion status	Index hospitalization (assessed up to 5 days)	2. OMI definitions using routine angiographic and laboratory parameters will be correlated to abnormal myocardial perfusion using correlation coefficients.
ECG prediction of mechanical reperfusion	Index hospitalization (assessed up to 5 days)	a. Correlation of AI-ECG TIMI model's raw numeric predictions (between 0 and 1) with myocardial perfusion grading post PCI.
Assessment of expert ECG interpretation blinded to all clinical information	Index hospitalization (assessed up to 5 days)	1. Accuracy, sensitivity, specificity, positive and negative predictive value of expert interpretation of active OMI, reperfused OMI and entirety of OMI; 2. Inter-rater variability of expert interpretation of active OMI, reperfused OMI and entirety of OMI.

Trial Locations

Locations (9)

LKH Hochsteiermark, Standort Bruck; 🇦🇹 Bruck an der Mur, Austria

LKH II Standort West; 🇦🇹 Graz, Austria

University Hospital D'Annunzio Chieti; 🇮🇹 Chieti, Italy

San Pietro Hospital; 🇮🇹 Rome, Italy

Sant´Andrea Hospital; 🇮🇹 Rome, Italy

Medical University of Graz; 🇦🇹 Graz, Austria

OLV Aalst; 🇧🇪 Aalst, Belgium

AZ Sint Jan; 🇧🇪 Brugge, Belgium

Jessa Ziekenhuis; 🇧🇪 Hasselt, Belgium