Can AI Detect What Is Not Injected? Evaluation of Lesion Detection in Virtual Contrast-Enhanced Breast MRI Using a Large-Scale AI Model Trained on GBCA-Enhanced Data
Tomography · 16.07.2026 · 12 (7) · 105
DOI: 10.3390/tomography12070105 ↗Kurz gefasst
Fachartikel in der Zeitschrift Tomography, online erschienen am 16.07.2026. Verfasst von Shirin Heidarikahkesh und 16 weiteren Autor:innen. Beteiligte Einrichtung: Institute of Radiology, Universitätsklinikum Erlangen, Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU), 91054 Erlangen, Germany. Hauptthemen laut MEDLINE: Contrast Media, Breast Neoplasms, Magnetic Resonance Imaging, Gadolinium. Die Arbeit nennt 49 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Bavarian Ministry of Economic Affairs and Media, Energy and Technology, Bavarian Ministry of Economic Affairs, Regional Development and Energy, German Research Foundation.
Abstract (Originalsprache)
Background/Objectives: Artificial intelligence (AI) can support lesion detection in gadolinium-based contrast agent-enhanced (GBCA-enhanced) breast MRI. However, its effectiveness on virtual contrast-enhanced (vCE) images remains unclear. This feasibility study evaluated the publicly available MAMA-MIA nnU-Net model trained on GBCA-enhanced data using an independent cohort of both GBCA-enhanced and vCE breast MRI.
Methods
This IRB-approved retrospective study included the publicly available nnU-Net model trained on n = 1506 MAMA-MIA breast MRI scans and a cohort of n = 2126 in-house 3T breast MRI scans. A generative adversarial network (Pix2Pix-GAN) was developed on n = 1870 of the in-house scans and used to generate vCE data on the remaining independent n = 256 in-house cases. The MAMA-MIA nnU-net was applied to both GBCA-enhanced (GBCA) and corresponding vCE images. Ground-truth segmentations of malignant lesions served to calculate the Dice score, Hausdorff distance, and lesion dimension differences.
Results
The final test set comprised n = 250 cases (n = 69 malignant, n = 181 benign). Lesion detection rates were 91% (n = 63/n = 69; 95% confidence interval (CI): 82.3–96.0%) for GBCA and 84% (n = 58/n = 69; 95% CI: 73.7–90.9%) for vCE. Two lesions missed in GBCA were identified by vCE. The Hausdorff distances were similar (GBCA: 6.4 (IQR: 3.2–9.3; 95% CI: 5.2–7.8) mm; vCE: 6.7 (IQR: 3.9–9.7; 95% CI: 5.3–8.0) mm, p = 0.564). The Dice scores showed minor differences (GBCA: 0.829 (IQR: 0.723–0.900; 95% CI: 0.786–0.865) vs. vCE: 0.826 (IQR: 0.720–0.857; 95% CI: 0.770–0.836); p < 0.001). vCE images had slightly higher non-target tissue segmentation (median 6072 mm3 vs. 5754 mm3).
Conclusions
A GBCA-trained algorithm demonstrated some cross-domain transferability to vCE images, albeit with a reduced case-level sensitivity of 84% (95% CI: 73.7–90.9%) vs. 91% (95% CI: 82.3–96.0%). Based on these preliminary results, further research, including larger cohorts and more diverse datasets, is warranted.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Substanzen
Autor:innen und Institutionen
- Institute of Radiology, Universitätsklinikum Erlangen, Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU), 91054 Erlangen, Germany.
- Pattern Recognition Lab., Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU), 91054 Erlangen, Germany.
- Department Artificial Intelligence in Biomedical Engineering, Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU), 91054 Erlangen, Germany.
- Department Medical Informatics, Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU), 91054 Erlangen, Germany.
- Institute of Computer Science, Polish Academy of Science, 01-248 Warsaw, Poland.
ORCID-Profile: Shirin Heidarikahkesh, Dominika Skwierawska, Luise Brock, Dominique Hadler, Frederik B Laun, Chris Ehring, Lorenz Döppmann, Ihor Horishnyi, Lorenz A Kapsner, Sabine Ohlmeyer, Andrzej Liebert
Förderung
- Bavarian Ministry of Economic Affairs and Media, Energy and Technology
- Bavarian Ministry of Economic Affairs, Regional Development and Energy
- German Research Foundation
- Bavarian State Ministry of Science
Zitieren
APA
Heidarikahkesh, S., Schreiter, H., George, A., Nguyen, T. T., Skwierawska, D., Brock, L., Hadler, D., Uder, M., Laun, F. B., Ehring, C., Graber, J., Döppmann, L., Horishnyi, I., Kapsner, L. A., Ohlmeyer, S., Liebert, A., & Bickelhaupt, S. (2026). Can AI Detect What Is Not Injected? Evaluation of Lesion Detection in Virtual Contrast-Enhanced Breast MRI Using a Large-Scale AI Model Trained on GBCA-Enhanced Data. Tomography, 12(7), 105. https://doi.org/10.3390/tomography12070105
Vancouver
Heidarikahkesh S, Schreiter H, George A, Nguyen TT, Skwierawska D, Brock L, et al. Can AI Detect What Is Not Injected? Evaluation of Lesion Detection in Virtual Contrast-Enhanced Breast MRI Using a Large-Scale AI Model Trained on GBCA-Enhanced Data. Tomography. 2026;12(7):105. doi:10.3390/tomography12070105
BibTeX
@article{Heidarikahkesh2026,
author = {Heidarikahkesh, Shirin and Schreiter, Hannes and George, Aju and Nguyen, Tri-Thien and Skwierawska, Dominika and Brock, Luise and Hadler, Dominique and Uder, Michael and Laun, Frederik B and Ehring, Chris and Graber, Johanna and Döppmann, Lorenz and Horishnyi, Ihor and Kapsner, Lorenz A and Ohlmeyer, Sabine and Liebert, Andrzej and Bickelhaupt, Sebastian},
title = {Can AI Detect What Is Not Injected? Evaluation of Lesion Detection in Virtual Contrast-Enhanced Breast MRI Using a Large-Scale AI Model Trained on GBCA-Enhanced Data},
journal = {Tomography},
year = {2026},
volume = {12},
number = {7},
pages = {105},
doi = {10.3390/tomography12070105}
}
EUVIMED