Designing a computer-assisted diagnosis system for cardiomegaly detection and radiology report generation
PLOS Digital Health · 20.05.2025 · 4 (5) · e0000835
DOI: 10.1371/journal.pdig.0000835 ↗Kurz gefasst
Fachartikel in der Zeitschrift PLOS Digital Health, online erschienen am 20.05.2025. Verfasst von Tianhao Zhu und 7 weiteren Autor:innen. Beteiligte Einrichtung: Section of Bioinformatics, Department of Metabolism, Digestion and Reproduction, Imperial College London, London, United Kingdom. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 56 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Medical Research Council, HORIZON EUROPE Framework Programme, UK Research and Innovation.
Abstract (Originalsprache)
Chest X-ray (CXR) is a diagnostic tool for cardiothoracic assessment. They make up 50% of all diagnostic imaging tests. With hundreds of images examined every day, radiologists can suffer from fatigue. This fatigue may reduce diagnostic accuracy and slow down report generation. We describe a prototype computer-assisted diagnosis (CAD) pipeline employing computer vision (CV) and Natural Language Processing (NLP). It was trained and evaluated on the publicly available MIMIC-CXR dataset. We perform image quality assessment, view labelling, and segmentation-based cardiomegaly severity classification. We use the output of the severity classification for large language model-based report generation. Four board-certified radiologists assessed the output accuracy of our CAD pipeline. Across the dataset composed of 377,100 CXR images and 227,827 free-text radiology reports, our system identified 0.18% of cases with mixed-sex mentions, 0.02% of poor quality images ( F 1 = 0.81), and 0.28% of wrongly labelled views (accuracy 99.4%). We assigned views for 4.18% of images which have unlabelled views. Our binary cardiomegaly classification model has 95.2% accuracy. The inter-radiologist agreement on evaluating the generated report’s semantics and correctness for radiologist-MIMIC is 0.62 (strict agreement) and 0.85 (relaxed agreement) similar to the radiologist-CAD agreement of 0.55 (strict) and 0.93 (relaxed). Our work found and corrected several incorrect or missing metadata annotations for the MIMIC-CXR dataset. The performance of our CAD system suggests performance on par with human radiologists. Future improvements revolve around improved text generation and the development of CV tools for other diseases.
Autor:innen und Institutionen
- Section of Bioinformatics, Department of Metabolism, Digestion and Reproduction, Imperial College London, London, United Kingdom.
- Department of Surgery and Cancer, Imperial College London, London, United Kingdom.
- Department of Computer Science, University College London, London, United Kingdom.
- Lysholm Department of Neuroradiology, National Hospital for Neurology and Neurosurgery, University College London Hospitals NHS Foundation Trust, London, United Kingdom.
ORCID-Profile: Kexin Xu, Matt Grech-Sollars, Joram M Posma
Förderung
- Medical Research Council
- HORIZON EUROPE Framework Programme
- UK Research and Innovation
Zitieren
APA
Zhu, T., Xu, K., Son, W., Linton-Reid, K., Boubnovski-Martell, M., Grech-Sollars, M., Lain, A. D., & Posma, J. M. (2025). Designing a computer-assisted diagnosis system for cardiomegaly detection and radiology report generation. PLOS Digital Health, 4(5), e0000835. https://doi.org/10.1371/journal.pdig.0000835
Vancouver
Zhu T, Xu K, Son W, Linton-Reid K, Boubnovski-Martell M, Grech-Sollars M, et al. Designing a computer-assisted diagnosis system for cardiomegaly detection and radiology report generation. PLOS Digital Health. 2025;4(5):e0000835. doi:10.1371/journal.pdig.0000835
BibTeX
@article{Zhu2025,
author = {Zhu, Tianhao and Xu, Kexin and Son, Wonchan and Linton-Reid, Kristofer and Boubnovski-Martell, Marc and Grech-Sollars, Matt and Lain, Antoine D and Posma, Joram M},
title = {Designing a computer-assisted diagnosis system for cardiomegaly detection and radiology report generation},
journal = {PLOS Digital Health},
year = {2025},
volume = {4},
number = {5},
pages = {e0000835},
doi = {10.1371/journal.pdig.0000835}
}
EUVIMED