Diagnostic accuracy of artificial intelligence-based models for acute myocardial infarction: a systematic review and meta-analysis
Diagnostische Genauigkeit KI-basierter Modelle für den akuten Myokardinfarkt: systematische Übersicht und Metaanalyse
Cardiovascular Research · 23.10.2026
DOI: 10.1093/cvr/cvag248 ↗Auf einen Blick
In 24 Studien waren KI-Modelle für die AMI-Diagnose sehr genau. Die Gesamtgüte lag bei AUC 0,97; für STEMI und NSTEMI blieben die Ergebnisse ähnlich.
Fragestellung
Wie genau sind KI-basierte Modelle zur Diagnose des akuten Myokardinfarkts insgesamt sowie getrennt für STEMI und NSTEMI?
Methode
Systematische Übersicht und Metaanalyse von Validierungs- und Testkohorten; hierarchische Modelle für AUC, Sensitivität und Spezifität; 24 Studien mit 27 Algorithmen.
Ergebnisse
166.052 Patient:innen; 18.850 (11,3 %) mit AMI. AUC 0,97, Sensitivität 0,93, Spezifität 0,92. STEMI: AUC 0,96, 0,89, 0,93. NSTEMI: AUC 0,97, 0,95, 0,90.
Einschränkungen
Direkte Head-to-Head-Vergleiche mit traditionellen statistischen Modellen waren laut Abstract begrenzt.
Schlussfolgerung der Autor:innen
KI-basierte diagnostische Algorithmen waren für die AMI-Diagnose besonders leistungsfähig und blieben auch für STEMI und NSTEMI konsistent. KI kann bestehende diagnostische Wege ergänzen.
Kernaussagen
- Gesamt-AUC 0,97 bei der AMI-Diagnose
- STEMI und NSTEMI ebenfalls mit hoher Genauigkeit
- Keine signifikanten Unterschiede nach KI-Typ oder Datenquelle
Deutsche Zusammenfassung: KI-gestützt aus dem Original-Abstract erstellt; alle Zahlen wurden automatisch mit dem Original abgeglichen. Maßgeblich ist der Original-Abstract.
Kurz gefasst
Systematische Übersicht in der Zeitschrift Cardiovascular Research, online erschienen am 23.10.2026. Verfasst von Federico Giacobbe und 25 weiteren Autor:innen. Beteiligte Einrichtung: Division of Cardiology, Cardiovascular and Thoracic Department, Città della Salute e della Scienza, Turin, Italy. Die Arbeit nennt 33 Literaturangaben.
Abstract (Originalsprache)
Aims
Despite continuous improvements in diagnostic tools, the diagnosis of acute myocardial infarction (AMI) is still based on electrocardiogram (ECG) and cardiac biomarkers supported by other non-invasive imaging techniques. However, ECG has limited sensitivity, while high-sensitivity troponin assays may show reduced specificity, particularly in older or multimorbid patients. Widely adopted clinical pathways (e.g. ESC 0/1 h algorithm) have reported high diagnostic discrimination in external cohorts (with AUC ranging from 0.93 to 0.96), which provides clinical context for interpreting our findings. These diagnostic tests have limitations that can be addressed through automated computer-based diagnostic systems using artificial intelligence (AI). Several studies in the literature have presented machine learning (ML) or deep learning (DL) models to quickly and accurately assess the diagnosis of AMI to optimize treatment. Methods and results The aim of our analysis was to evaluate the performance of AI-based models for the diagnosis of AMI using data from all validation and test cohorts available in the literature. Sensitivity and specificity for all MIs, and for ST-elevation myocardial infarction (STEMI) and non ST-elevation myocardial infarction (NSTEMI) separately, were analysed for different types of AI approaches (non-DL or DL) and for different data sources (ECG or troponin-based). Hierarchical models were used to estimate the summary receiver operating characteristic curve, sensitivity, and specificity. A total of 24 studies evaluating 27 AI-based diagnostic algorithms were included, encompassing 166.052 patients, of which 18.850 (11.3%) had a final diagnosis of AMI. Patients were predominantly male (61.8%) with a median age of 60.7 (IQR 59.0–64.1) years. Discrimination of AI-based algorithms for the diagnosis of AMI was high, with an AUC of 0.97 (sensitivity 0.93, specificity 0.92). This result remained consistent when considering the ability to diagnose STEMI (AUC 0.96, sensitivity 0.89, specificity 0.93) or NSTEMI (AUC 0.97, sensitivity 0.95, specificity 0.90) separately. No significant difference was found regarding the type of AI. ECG and troponin-based models were both extremely accurate, with the first ones being less sensitive and the latter less specific.
Conclusion
AI-based diagnostic algorithms for AMI diagnosis were associated with particularly efficient performance. These results were consistent for both STEMI and NSTEMI considered separately. Rather than replacing conventional diagnostic pathways, AI has the potential to enhance existing tools by identifying subtle ECG features and integrating multiple data streams. While direct head-to-head comparisons with traditional statistical models are currently limited, future prospective studies evaluating AI and conventional approaches within the same clinical populations are warranted. AI did not lose accuracy depending on the data source or the type of AI applied.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Division of Cardiology, Cardiovascular and Thoracic Department, Città della Salute e della Scienza, Turin, Italy.
- Division of Cardiology, Department of Medical Sciences, University of Turin, Italy.
- Royal Brompton and Harefield Hospitals Trust and Imperial College, London, UK.
- Seoul National University Bundang Hospital, CEO, ARPI Inc, Seongnam, Republic of Korea.
- School of Medicine, National Defense Medical Center, Taipei, Taiwan, Republic of China. No.161, Min-Chun E. Rd., Sec. 6, Neihu, Taipei 114, Taiwan, Republic of China.
- Digital Healthcare Institute, Sejong Medical Research Institute, Bucheon, Republic of Korea.
ORCID-Profile: Ovidio De Filippo, Joonghee Kim, Chin Lin, Joon-Myoung Kwon, Francesco Bruno, Kuan Ken Lee, Pedro Lopez Ayala, Nicholas Mills, Jasper Boeddinghaus, Christian Mueller, Piero Fariselli, Gaetano Maria De Ferrari
Zitieren
APA
Giacobbe, F., De Filippo, O., Petracca, M., Nebiolo, M., Cimino, R. A., Michelone, M., Capoccia, S., Kim, J., Lin, C., Kwon, J. M., Bruno, F., Balducci, M., Yoon, D., Lee, H. S., Doudesis, D., Lee, K. K., Ayala, P. L., Mills, N., Boeddinghaus, J., …, & D'Ascenzo, F. (2026). Diagnostic accuracy of artificial intelligence-based models for acute myocardial infarction: a systematic review and meta-analysis. Cardiovascular Research. https://doi.org/10.1093/cvr/cvag248
Vancouver
Giacobbe F, De Filippo O, Petracca M, Nebiolo M, Cimino RA, Michelone M, et al. Diagnostic accuracy of artificial intelligence-based models for acute myocardial infarction: a systematic review and meta-analysis. Cardiovascular Research. 2026. doi:10.1093/cvr/cvag248
BibTeX
@article{Giacobbe2026,
author = {Giacobbe, Federico and De Filippo, Ovidio and Petracca, Manuel and Nebiolo, Marco and Cimino, Rino Andrea and Michelone, Matteo and Capoccia, Sergio and Kim, Joonghee and Lin, Chin and Kwon, Joon-Myoung and Bruno, Francesco and Balducci, Marco and Yoon, Dukyong and Lee, Hak Seung and Doudesis, Dimitrios and Lee, Kuan Ken and Ayala, Pedro Lopez and Mills, Nicholas and Boeddinghaus, Jasper and Grangetto, Marco and Deriu, Marco Agostino and Mueller, Christian and Fariselli, Piero and Luscher, Thomas and De Ferrari, Gaetano Maria and D'Ascenzo, Fabrizio},
title = {Diagnostic accuracy of artificial intelligence-based models for acute myocardial infarction: a systematic review and meta-analysis},
journal = {Cardiovascular Research},
year = {2026},
doi = {10.1093/cvr/cvag248}
}
EUVIMED