SegmentAnyBone: A universal model that segments any bone at any location on MRI

• Published in: Medical image analysis Vol. 101; p. 103469
• Main Authors: Gu, Hanxue, Colglazier, Roy, Dong, Haoyu, Zhang, Jikai, Chen, Yaqian, Yildiz, Zafer, Chen, Yuwen, Li, Lin, Yang, Jichen, Willhite, Jay, Meyer, Alex M., Guo, Brian, Shah, Yashvi Atul, Luo, Emily, Rajput, Shipra, Kuehn, Sally, Bulleit, Clark, Wu, Kevin A., Lee, Jisoo, Ramirez, Brandon, Lu, Darui, Levin, Jay M., Mazurowski, Maciej A.
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 01.04.2025
• Subjects: Algorithms; Bone and Bones - diagnostic imaging; Bone segmentation; Deep Learning; Foundation model; Humans; Image Interpretation, Computer-Assisted - methods; Image Processing, Computer-Assisted - methods; Magnetic Resonance Imaging - methods; MRI; Deep learning; Bone segmentation; MRI; Foundation model
• ISSN: 1361-8415; 1361-8423; 1361-8423
• DOI: 10.1016/j.media.2025.103469

Magnetic Resonance Imaging (MRI) is pivotal in radiology, offering non-invasive and high-quality insights into the human body. Precise segmentation of the MRIs into different organs and tissues would be very beneficial as it would allow more accurate measurements, which are essential for accurate diagnosis and effective treatment planning. Specifically, segmenting bones in MRI would allow for more quantitative assessments of musculoskeletal conditions, while such assessments are largely absent in current radiological practice. The difficulty of bone MRI segmentation is illustrated by the fact that limited algorithms are publicly available, and those contained in the literature typically address a specific anatomic area. In our study, we propose a versatile, publicly available deep learning model for bone segmentation in MRI at multiple standard MRI locations. The proposed model can operate in two modes: fully automated segmentation and prompt-based segmentation. Our contributions include (1) collecting and annotating a new MRI dataset across various MRI protocols, encompassing 320 annotated volumes and more than 10k annotated slices across diverse anatomic regions; (2) investigating several standard network architectures and strategies for automated segmentation; (3) introducing SegmentAnyBone, an innovative foundation model-based approach that extends the Segment Anything Model (SAM); (4) comparative analysis of our algorithm and previous approaches; and (5) generalization analysis of our algorithm across different anatomical locations and MRI sequences, as well as three external datasets. We publicly release our model at Github Code. [Display omitted] •Launched SegmentAnyBone, a publicly accessible, versatile deep-learning model for MRI bone segmentation.•Adaptable to various standard MRI locations and operational in both fully automated and prompt-based modes.•Compiled and annotated a new MRI dataset with over 300 volumes and 10k slices across diverse anatomical regions.•Extended SAM with a foundation model-based approach and introduced a depth attention branch for 3D segmentation.•Validated superior performance and generalization across anatomy, MRI sequences, and three external datasets.•Released SegmentAnyBone codes, model, and fine-tuning scripts as a public resource for MRI bone segmentation research.