Cross-site Validation of AI Segmentation and Harmonization in Breast MRI Journal Article
| Authors: | Huang, Y.; Leotta, N. J.; Hirsch, L.; Lo Gullo, R.; Hughes, M.; Reiner, J.; Saphier, N. B.; Myers, K. S.; Panigrahi, B.; Ambinder, E.; Di Carlo, P.; Grimm, L. J.; Lowell, D.; Yoon, S.; Ghate, S. V.; Parra, L. C.; Sutton, E. J. |
| Article Title: | Cross-site Validation of AI Segmentation and Harmonization in Breast MRI |
| Abstract: | This work aims to perform a cross-site validation of automated segmentation for breast cancers in MRI and to compare the performance to radiologists. A three-dimensional (3D) U-Net was trained to segment cancers in dynamic contrast-enhanced axial MRIs using a large dataset from Site 1 (n = 15,266; 449 malignant and 14,817 benign). Performance was validated on site-specific test data from this and two additional sites, and common publicly available testing data. Four radiologists from each of the three clinical sites provided two-dimensional (2D) segmentations as ground truth. Segmentation performance did not differ between the network and radiologists on the test data from Sites 1 and 2 or the common public data (median Dice score Site 1, network 0.86 vs. radiologist 0.85, n = 114; Site 2, 0.91 vs. 0.91, n = 50; common: 0.93 vs. 0.90). For Site 3, an affine input layer was fine-tuned using segmentation labels, resulting in comparable performance between the network and radiologist (0.88 vs. 0.89, n = 42). Radiologist performance differed on the common test data, and the network numerically outperformed 11 of the 12 radiologists (median Dice: 0.85-0.94, n = 20). In conclusion, a deep network with a novel supervised harmonization technique matches radiologists' performance in MRI tumor segmentation across clinical sites. We make code and weights publicly available to promote reproducible AI in radiology. |
| Keywords: | mri; harmonization; dynamic contrast enhancement; cancer; deep learning; cross-site evaluation; breast cancer segmentation |
| Journal Title: | Journal of Imaging Informatics in Medicine |
| Volume: | 38 |
| Issue: | 3 |
| ISSN: | 2948-2925 |
| Publisher: | Springer |
| Date Published: | 2025-01-01 |
| Start Page: | 1642 |
| End Page: | 1652 |
| Language: | English |
| ACCESSION: | WOS:001320264000004 |
| DOI: | 10.1007/s10278-024-01266-9 |
| PROVIDER: | wos |
| PMCID: | PMC12092851 |
| PUBMED: | 39320547 |
| Notes: | Article -- Source: Wos |
Altmetric
Citation Impact
BMJ Impact Analytics
Related MSK Work