EUVIMEDEUVIMED entdecken
VergleichsstudieOpen Access · CC BY 4.0

Etiologic classification of suspected MINOCA using cardiovascular magnetic resonance reports: a comparison of a large language model and human readers

Mihály Károlyi, Verena C Wilzeck, Lucas Tramèr, Léon Groenhoff, Jochen von Spiczak, Tamar Bigvava, Hatem Alkadhi, Robert Manka

The International Journal of Cardiovascular Imaging · 2026

DOI: 10.1007/s10554-026-03689-7 ↗

Kurz gefasst

Vergleichsstudie in der Zeitschrift The International Journal of Cardiovascular Imaging, 2026. Verfasst von Mihály Károlyi und 7 weiteren Autor:innen. Beteiligte Einrichtung: Diagnostic and Interventional Radiology, University Hospital Zurich, University of Zurich, Zurich, Switzerland. Hauptthemen laut MEDLINE: Large Language Models, Myocardial Infarction, Coronary Artery Disease, Decision Support Techniques. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 28 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The study explored the feasibility of using a large language model (LLM) for etiologic classification in patients with suspected myocardial infarction with non-obstructive coronary arteries (MINOCA) based on cardiovascular magnetic resonance (CMR) reports. We included 156 patients with MINOCA from a prospective (n = 50) and a retrospective (n = 106) pooled cohort. A large language model and three human readers with different experience levels independently classified CMR reports into eight predefined diagnostic categories, using the final expert CMR diagnosis as the reference standard. Performance and agreement were assessed using standard multiclass classification metrics. In the pooled cohort, the LLM achieved an exact-match diagnostic accuracy of 67.3% (95% CI 59.6-74.2%), lower than the expert reader (80.1%, 95% CI 73.2-85.6) but comparable to the intermediate and junior readers (both 68.6%, 95% CI 60.9-75.4%). Diagnosis-specific analysis showed consistently high specificity for the LLM (mean 94.9%), with sensitivities up to 86.0% and F1-scores up to 85.1% for common etiologies. Agreement between the LLM and the reference standard was substantial (ICC 0.84, 95% CI: 0.79-0.89) and comparable to experienced readers, whereas agreement with the junior reader was markedly lower, indicating greater diagnostic variability despite similar accuracy. A large language model shows promise as a supportive tool for etiologic classification in suspected MINOCA from CMR reports, with performance comparable to less experienced readers and high agreement with the final expert CMR diagnosis. Integration into structured reporting workflows may enhance diagnostic consistency in routine clinical practice.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Large Language ModelsMyocardial InfarctionCoronary Artery DiseaseDecision Support TechniquesMagnetic Resonance ImagingRadiologistsHumansPredictive Value of TestsFemaleMaleReproducibility of ResultsObserver VariationFeasibility StudiesRetrospective StudiesMiddle AgedProspective StudiesAgedClinical CompetenceImage Interpretation, Computer-Assisted

Stichwörter der Autor:innen

Diagnostic classificationLarge language modelMINOCAMagnetic resonance imaging

Autor:innen und Institutionen

  • Diagnostic and Interventional Radiology, University Hospital Zurich, University of Zurich, Zurich, Switzerland.
  • Department of Cardiology, University Heart Center, University Hospital Zurich, University of Zurich, Raemistrasse 100, Zurich, 8091, Switzerland.
  • Institute for Biomedical Engineering, University and ETH Zurich, Zurich, Switzerland.
  • Diagnostic and Interventional Radiology, University Hospital Zurich, University of Zurich, Zurich, Switzerland. Robert.Manka@usz.ch.
  • Department of Cardiology, University Heart Center, University Hospital Zurich, University of Zurich, Raemistrasse 100, Zurich, 8091, Switzerland. Robert.Manka@usz.ch.
  • Institute for Biomedical Engineering, University and ETH Zurich, Zurich, Switzerland. Robert.Manka@usz.ch.

Zitieren

APA

Károlyi, M., Wilzeck, V. C., Tramèr, L., Groenhoff, L., von Spiczak, J., Bigvava, T., Alkadhi, H., & Manka, R. (2026). Etiologic classification of suspected MINOCA using cardiovascular magnetic resonance reports: a comparison of a large language model and human readers. The International Journal of Cardiovascular Imaging. https://doi.org/10.1007/s10554-026-03689-7

Vancouver

Károlyi M, Wilzeck VC, Tramèr L, Groenhoff L, von Spiczak J, Bigvava T, et al. Etiologic classification of suspected MINOCA using cardiovascular magnetic resonance reports: a comparison of a large language model and human readers. The International Journal of Cardiovascular Imaging. 2026. doi:10.1007/s10554-026-03689-7

BibTeX
@article{Krolyi2026,
  author = {Károlyi, Mihály and Wilzeck, Verena C and Tramèr, Lucas and Groenhoff, Léon and von Spiczak, Jochen and Bigvava, Tamar and Alkadhi, Hatem and Manka, Robert},
  title = {Etiologic classification of suspected MINOCA using cardiovascular magnetic resonance reports: a comparison of a large language model and human readers},
  journal = {The International Journal of Cardiovascular Imaging},
  year = {2026},
  doi = {10.1007/s10554-026-03689-7}
}
Kostenlos starten →