International multicenter validation of AI-driven ultrasound detection of ovarian cancer
Nature Medicine · 2025-01 · 31 (1) · 189-196
DOI: 10.1038/s41591-024-03329-4 ↗Kurz gefasst
Multizentrische Studie in der Zeitschrift Nature Medicine, online erschienen am 02.01.2025. Verfasst von Filip Christiansen und 30 weiteren Autor:innen. Beteiligte Einrichtung: Department of Clinical Science and Education, Södersjukhuset, Karolinska Institutet, Stockholm, Sweden. Hauptthemen laut MEDLINE: Ovarian Neoplasms, Artificial Intelligence. Die Arbeit wurde bisher 70-mal zitiert (Crossref) und nennt 50 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Radiumhemmets Forskningsfonder (Cancer Research Foundations of Radiumhemmet), Cancerfonden (Swedish Cancer Society), Vetenskapsrdet (Swedish Research Council).
Abstract (Originalsprache)
Ovarian lesions are common and often incidentally detected. A critical shortage of expert ultrasound examiners has raised concerns of unnecessary interventions and delayed cancer diagnoses. Deep learning has shown promising results in the detection of ovarian cancer in ultrasound images; however, external validation is lacking. In this international multicenter retrospective study, we developed and validated transformer-based neural network models using a comprehensive dataset of 17,119 ultrasound images from 3,652 patients across 20 centers in eight countries. Using a leave-one-center-out cross-validation scheme, for each center in turn, we trained a model using data from the remaining centers. The models demonstrated robust performance across centers, ultrasound systems, histological diagnoses and patient age groups, significantly outperforming both expert and non-expert examiners on all evaluated metrics, namely F1 score, sensitivity, specificity, accuracy, Cohen’s kappa, Matthew’s correlation coefficient, diagnostic odds ratio and Youden’s J statistic. Furthermore, in a retrospective triage simulation, artificial intelligence (AI)-driven diagnostic support reduced referrals to experts by 63% while significantly surpassing the diagnostic performance of the current practice. These results show that transformer-based models exhibit strong generalization and above human expert-level diagnostic accuracy, with the potential to alleviate the shortage of expert ultrasound examiners and improve patient outcomes.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Department of Clinical Science and Education, Södersjukhuset, Karolinska Institutet, Stockholm, Sweden.
- Department of Obstetrics and Gynecology, Södersjukhuset, Stockholm, Sweden.
- School of Electrical Engineering and Computer Science, KTH Royal Institute of Technology, Stockholm, Sweden.
- Science for Life Laboratory, Stockholm, Sweden.
- Department of Gynecological Oncology and Gynecology, Medical University of Lublin, Lublin, Poland.
- Unit of Obstetrics & Gynecology, Department of Biomedical and Clinical Sciences, Luigi Sacco University Hospital, University of Milan, Milan, Italy.
ORCID-Profile: Filip Christiansen, Emir Konuk, Adithya Raju Ganeshan, Robert Welch, Joana Palés Huix, Lucia Anna Haak, Robert Fruscio, Dorella Franchi, Daniela Fischerova, Maria Àngela Pascual, Francesca Buonomo, Ekaterini Domali, Chiara Carella, Petra Saskova, Debora Verri, Elisabeth Epstein
Förderung
- Radiumhemmets Forskningsfonder (Cancer Research Foundations of Radiumhemmet)
- Cancerfonden (Swedish Cancer Society)
- Vetenskapsrdet (Swedish Research Council)
Zitieren
APA
Christiansen, F., Konuk, E., Ganeshan, A. R., Welch, R., Palés Huix, J., Czekierdowski, A., Leone, F. P. G., Haak, L. A., Fruscio, R., Gaurilcikas, A., Franchi, D., Fischerova, D., Mor, E., Savelli, L., Pascual, M. À., Kudla, M. J., Guerriero, S., Buonomo, F., Liuba, K., …, & Epstein, E. (2025). International multicenter validation of AI-driven ultrasound detection of ovarian cancer. Nature Medicine, 31(1), 189-196. https://doi.org/10.1038/s41591-024-03329-4
Vancouver
Christiansen F, Konuk E, Ganeshan AR, Welch R, Palés Huix J, Czekierdowski A, et al. International multicenter validation of AI-driven ultrasound detection of ovarian cancer. Nature Medicine. 2025;31(1):189-196. doi:10.1038/s41591-024-03329-4
BibTeX
@article{Christiansen2025,
author = {Christiansen, Filip and Konuk, Emir and Ganeshan, Adithya Raju and Welch, Robert and Palés Huix, Joana and Czekierdowski, Artur and Leone, Francesco Paolo Giuseppe and Haak, Lucia Anna and Fruscio, Robert and Gaurilcikas, Adrius and Franchi, Dorella and Fischerova, Daniela and Mor, Elisa and Savelli, Luca and Pascual, Maria Àngela and Kudla, Marek Jerzy and Guerriero, Stefano and Buonomo, Francesca and Liuba, Karina and Montik, Nina and Alcázar, Juan Luis and Domali, Ekaterini and Pangilinan, Nelinda Catherine P and Carella, Chiara and Munaretto, Maria and Saskova, Petra and Verri, Debora and Visenzi, Chiara and Herman, Pawel and Smith, Kevin and Epstein, Elisabeth},
title = {International multicenter validation of AI-driven ultrasound detection of ovarian cancer},
journal = {Nature Medicine},
year = {2025},
volume = {31},
number = {1},
pages = {189--196},
doi = {10.1038/s41591-024-03329-4}
}
EUVIMED