Multidisciplinary Consensus Prostate Contours on Magnetic Resonance Imaging: Educational Atlas and Reference Standard for Artificial Intelligence Benchmarking
International Journal of Radiation Oncology*Biology*Physics · 2025
DOI: 10.1016/j.ijrobp.2025.03.024 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Radiation Oncology*Biology*Physics, 2025. Verfasst von Yuze Song und 19 weiteren Autor:innen. Beteiligte Einrichtung: Department of Radiation Medicine and Applied Sciences, University of California San Diego, La Jolla, California; Department of Electrical and Compute…. Hauptthemen laut MEDLINE: Magnetic Resonance Imaging, Artificial Intelligence, Prostatic Neoplasms, Benchmarking. Die Arbeit wurde bisher 13-mal zitiert (Crossref) und nennt 36 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: NIBIB NIH HHS, NCI NIH HHS, NCATS NIH HHS.
Abstract (Originalsprache)
PURPOSE: Evaluation of artificial intelligence (AI) algorithms for prostate segmentation is challenging because ground truth is lacking. We aimed to: (1) create a reference standard data set with precise prostate contours by expert consensus, and (2) evaluate various AI tools against this standard. METHODS AND MATERIALS: We obtained prostate magnetic resonance imaging cases from six institutions from the Qualitative Prostate Imaging Consortium. A panel of 4 experts (2 genitourinary radiologists and 2 prostate radiation oncologists) meticulously developed consensus prostate segmentations on axial T2-weighted series. We evaluated the performance of 6 AI tools (3 commercially available and 3 academic) using Dice scores, distance from reference contour, and volume error. RESULTS: The panel achieved consensus prostate segmentation on each slice of all 68 patient cases included in the reference data set. We present 2 patient examples to serve as contouring guides. Depending on the AI tool, median Dice scores (across patients) ranged from 0.80 to 0.94 for whole prostate segmentation. For a typical (median) patient, AI tools had a mean error over the prostate surface ranging from 1.3 to 2.4 mm. They maximally deviated 3.0 to 9.4 mm outside the prostate and 3.0 to 8.5 mm inside the prostate for a typical patient. Error in prostate volume measurement for a typical patient ranged from 4.3% to 31.4%. CONCLUSIONS: We established an expert consensus benchmark for prostate segmentation. The best-performing AI tools have typical accuracy greater than that reported for radiation oncologists using computed tomography scans (the most common clinical approach for radiation therapy planning). Physician review remains essential to detect occasional major errors.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Department of Radiation Medicine and Applied Sciences, University of California San Diego, La Jolla, California; Department of Electrical and Computer Engineering, University of California San Diego, La Jolla, Californi…
- Department of Radiation Medicine and Applied Sciences, University of California San Diego, La Jolla, California.
- Department of Radiation Oncology, University of Michigan, Ann Arbor, Michigan.
- Department of Radiology, Cornell University, Ithaca, New York.
- Department of Clinical Imaging Sciences, University of Rochester Medical Center, Rochester, New York.
- Department of Radiology, University of Cambridge, Cambridge, United Kingdom.
Förderung
- NIBIB NIH HHS
- NCI NIH HHS
- NCATS NIH HHS
- Siemens
- National Institute of Biomedical Imaging and Bioengineering
- The Slovenian Research and Innovation Agency
- American Society for Radiation Oncology
- National Institutes of Health
Zitieren
APA
Song, Y., Dornisch, A. M., Dess, R. T., Margolis, D. J. A., Weinberg, E. P., Barrett, T., Cornell, M., Fan, R. E., Harisinghani, M., Kamran, S. C., Lee, J. H., Li, C. X., Liss, M. A., Rusu, M., Santos, J., Sonn, G. A., Vidic, I., Woolen, S. A., Dale, A. M., & Seibert, T. M. (2025). Multidisciplinary Consensus Prostate Contours on Magnetic Resonance Imaging: Educational Atlas and Reference Standard for Artificial Intelligence Benchmarking. International Journal of Radiation Oncology*Biology*Physics. https://doi.org/10.1016/j.ijrobp.2025.03.024
Vancouver
Song Y, Dornisch AM, Dess RT, Margolis DJA, Weinberg EP, Barrett T, et al. Multidisciplinary Consensus Prostate Contours on Magnetic Resonance Imaging: Educational Atlas and Reference Standard for Artificial Intelligence Benchmarking. International Journal of Radiation Oncology*Biology*Physics. 2025. doi:10.1016/j.ijrobp.2025.03.024
BibTeX
@article{Song2025,
author = {Song, Yuze and Dornisch, Anna M and Dess, Robert T and Margolis, Daniel J A and Weinberg, Eric P and Barrett, Tristan and Cornell, Mariel and Fan, Richard E and Harisinghani, Mukesh and Kamran, Sophia C and Lee, Jeong Hoon and Li, Cynthia Xinran and Liss, Michael A and Rusu, Mirabela and Santos, Jason and Sonn, Geoffrey A and Vidic, Igor and Woolen, Sean A and Dale, Anders M and Seibert, Tyler M},
title = {Multidisciplinary Consensus Prostate Contours on Magnetic Resonance Imaging: Educational Atlas and Reference Standard for Artificial Intelligence Benchmarking},
journal = {International Journal of Radiation Oncology*Biology*Physics},
year = {2025},
doi = {10.1016/j.ijrobp.2025.03.024}
}
EUVIMED