Deep Learning-Based Detection and Classification of Bone Lesions on Staging Computed Tomography in Prostate Cancer: A Development Study

Mason J Belue; Stephanie A Harmon; Dong Yang; Julie Y An; Sonia Gaur; Yan Mee Law; Evrim Turkbey; Ziyue Xu; Jesse Tetreault; Nathan S Lay; Enis C Yilmaz; Tim E Phelps; Benjamin Simon; Liza Lindenberg; Esther Mena; Peter A Pinto; Ulas Bagci; Bradford J Wood; Deborah E Citrin; William L Dahut; Ravi A Madan; James L Gulley; Daguang Xu; Peter L Choyke; Baris Turkbey

doi:10.1016/j.acra.2024.01.009

Deep Learning-Based Detection and Classification of Bone Lesions on Staging Computed Tomography in Prostate Cancer: A Development Study

Acad Radiol. 2024 Jan 22:S1076-6332(24)00008-4. doi: 10.1016/j.acra.2024.01.009. Online ahead of print.

Authors

Mason J Belue¹, Stephanie A Harmon¹, Dong Yang², Julie Y An³, Sonia Gaur⁴, Yan Mee Law⁵, Evrim Turkbey⁶, Ziyue Xu², Jesse Tetreault², Nathan S Lay¹, Enis C Yilmaz¹, Tim E Phelps¹, Benjamin Simon¹, Liza Lindenberg¹, Esther Mena¹, Peter A Pinto⁷, Ulas Bagci⁸, Bradford J Wood⁹, Deborah E Citrin¹⁰, William L Dahut¹¹, Ravi A Madan¹¹, James L Gulley¹², Daguang Xu², Peter L Choyke¹, Baris Turkbey¹³

Affiliations

¹ Molecular Imaging Branch, National Cancer Institute, National Institutes of Health, 10 Center Dr., MSC 1182, Building 10, Room B3B85, Bethesda, Maryland, USA (M.J.B., S.A.H., N.S.L., E.C.Y., T.E.P., B.S., L.L., E.M., P.L.C., B.T.).
² NVIDIA Corporation, Santa Clara, California, USA (D.Y., Z.X., J.T., D.X.).
³ Department of Radiology, University of California, San Diego, California, USA (J.Y.A.).
⁴ Department of Radiology, Massachusetts General Hospital, Boston, Massachusetts, USA (S.G.).
⁵ Department of Radiology, Singapore General Hospital, Singapore (Y.M.L.).
⁶ Department of Radiology, Clinical Center, National Institutes of Health, Bethesda, Maryland, USA (E.T., B.J.W.).
⁷ Urologic Oncology Branch, National Cancer Institute, National Institutes of Health, Bethesda, Maryland, USA (P.A.P.).
⁸ Radiology and Biomedical Engineering Department, Northwestern University Feinberg School of Medicine, Chicago, Illinois, USA (U.B.).
⁹ Department of Radiology, Clinical Center, National Institutes of Health, Bethesda, Maryland, USA (E.T., B.J.W.); Center for Interventional Oncology, National Cancer Institute, National Institutes of Health, Bethesda, Maryland, USA (B.J.W.).
¹⁰ Radiation Oncology Branch, National Cancer Institute, National Institutes of Health, Bethesda, Maryland, USA (D.E.C.).
¹¹ Genitourinary Malignancies Branch, National Cancer Institute, National Institutes of Health, Bethesda, Maryland, USA (W.L.D., R.A.M.).
¹² Center for Immuno-Oncology, National Cancer Institute, National Institutes of Health, Bethesda, Maryland, USA (J.L.G.).
¹³ Molecular Imaging Branch, National Cancer Institute, National Institutes of Health, 10 Center Dr., MSC 1182, Building 10, Room B3B85, Bethesda, Maryland, USA (M.J.B., S.A.H., N.S.L., E.C.Y., T.E.P., B.S., L.L., E.M., P.L.C., B.T.). Electronic address: turkbeyi@mail.nih.gov.

PMID: 38262813
DOI: 10.1016/j.acra.2024.01.009

Abstract

Rationale and objectives: Efficiently detecting and characterizing metastatic bone lesions on staging CT is crucial for prostate cancer (PCa) care. However, it demands significant expert time and additional imaging such as PET/CT. We aimed to develop an ensemble of two automated deep learning AI models for 1) bone lesion detection and segmentation and 2) benign vs. metastatic lesion classification on staging CTs and to compare its performance with radiologists.

Materials and methods: This retrospective study developed two AI models using 297 staging CT scans (81 metastatic) with 4601 benign and 1911 metastatic lesions in PCa patients. Metastases were validated by follow-up scans, bone biopsy, or PET/CT. Segmentation AI (3DAISeg) was developed using the lesion contours delineated by a radiologist. 3DAISeg performance was evaluated with the Dice similarity coefficient, and classification AI (3DAIClass) performance on AI and radiologist contours was assessed with F1-score and accuracy. Training/validation/testing data partitions of 70:15:15 were used. A multi-reader study was performed with two junior and two senior radiologists within a subset of the testing dataset (n = 36).

Results: In 45 unseen staging CT scans (12 metastatic PCa) with 669 benign and 364 metastatic lesions, 3DAISeg detected 73.1% of metastatic (266/364) and 72.4% of benign lesions (484/669). Each scan averaged 12 extra segmentations (range: 1-31). All metastatic scans had at least one detected metastatic lesion, achieving a 100% patient-level detection. The mean Dice score for 3DAISeg was 0.53 (median: 0.59, range: 0-0.87). The F1 for 3DAIClass was 94.8% (radiologist contours) and 92.4% (3DAISeg contours), with a median false positive of 0 (range: 0-3). Using radiologist contours, 3DAIClass had PPV and NPV rates comparable to junior and senior radiologists: PPV (semi-automated approach AI 40.0% vs. Juniors 32.0% vs. Seniors 50.0%) and NPV (AI 96.2% vs. Juniors 95.7% vs. Seniors 91.9%). When using 3DAISeg, 3DAIClass mimicked junior radiologists in PPV (pure-AI 20.0% vs. Juniors 32.0% vs. Seniors 50.0%) but surpassed seniors in NPV (pure-AI 93.8% vs. Juniors 95.7% vs. Seniors 91.9%).

Conclusion: Our lesion detection and classification AI model performs on par with junior and senior radiologists in discerning benign and metastatic lesions on staging CTs obtained for PCa.

Keywords: Artificial intelligence; Bone metastasis; Computed tomography; Deep learning; MONAI; Oligometastatic; Prostate cancer.

Published by Elsevier Inc.