Key result
AI-enabled ECG detects acute MI with ~89% sensitivity and high overall diagnostic accuracy.
Why the study?
Does AI-enabled ECG accurately detect acute myocardial infarction (including STEMI and NSTEMI) in patients undergoing clinical triage?
Meta-Analysis (n=94,510)
Yes
Does AI-enabled ECG accurately detect acute myocardial infarction (including STEMI and NSTEMI) in patients undergoing clinical triage?
AI-enabled ECG demonstrates high diagnostic accuracy for overall AMI and STEMI detection (AUC 0.97 and 0.98, respectively) to support adjunctive clinical triage, though lower sensitivity for NSTEMI (65.0%) precludes its use as a standalone rule-out test.
High diagnostic accuracy supports AI-ECG evaluation for acute MI; leaves open integration into clinical workflows without randomized evidence.
No abstract yet. The journal hasn't published one we can show.
No takes yet. Share an insight, caveat, or question.
Gaelzer et al. (2026) conducted a meta-analysis in Acute myocardial infarction (AMI) (n=94,510). AI-enabled ECG (AI-ECG) vs. Reference standard diagnosis of acute myocardial infarction was evaluated on Diagnostic accuracy for overall acute myocardial infarction detection (AUC 0.97 (sensitivity 89.4%, specificity 96.0%)). AI-enabled ECG detected overall acute myocardial infarction with a pooled sensitivity of 89.4%, specificity of 96.0%, and an area under the curve of 0.97 across 94,510 participants.
Synapse has enriched 5 related papers. Consider them for comparative context: