Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Lian Zhang1, Zhengliang Liu2, Lu Zhang3

  • 1Department of Radiation Oncology, Mayo Clinic, Phoenix, Arizona, USA.

Medical Physics
|February 6, 2024
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Research through Evaluation for Large Language Model in Patient-Clinician Communications.

Research square·2026
Same author

Community-level modeling of gyral folding patterns for robust and anatomically informed individualized brain mapping.

NeuroImage·2026
Same author

AD-GPT: large language models in Alzheimer's disease.

BMC medical informatics and decision making·2026
Same author

Noisy probing dose facilitated dose prediction for pencil beam scanning proton therapy: Physics enhances generalizability.

Medical physics·2026
Same author

Large language models for bioinformatics.

Quantitative biology (Beijing, China)·2026
Same author

Biophysical modeling of anatomically realistic prenatal cortical folding development.

Research square·2026
Same journal

Correction to "On the shape of the radiation survival curve in tumor spheroids: The role of oxygen heterogeneity".

Medical physics·2026
Same journal

Multi-view constrained semi-supervised vertebra detection for 3D ultrasound spine volume.

Medical physics·2026
Same journal

Accuracy of quantitative <sup>177</sup>Lu SPECT/CT imaging: A systematic review.

Medical physics·2026
Same journal

Physics-constrained dual-domain network for CBCT reconstruction from orthogonal X-rays in gynecologic radiotherapy.

Medical physics·2026
Same journal

Decomposition-based harmonization for quantitative PET imaging across scanners and radiotracers.

Medical physics·2026
Same journal

Development and evaluation of an in vivo dose-based monitoring system for electron FLASH radiation therapy.

Medical physics·2026
See all related articles

Meta's Segment Anything Model (SAM) shows promise for automatic organ at risk segmentation in radiation therapy, improving efficiency with interactive prompting for patient-specific plans.

Area of Science:

  • Radiotherapy and Medical Imaging

Context:

  • Accurate delineation of organs at risk (OARs) is crucial for radiation therapy planning and dose evaluation.
  • Deep learning models offer potential for auto-segmentation but face challenges in generalizability and human-AI interaction.
  • A generalizable and promptable model is needed for efficient, multi-site OAR segmentation with interactive capabilities.

Purpose:

  • To evaluate the performance of Meta's Segment Anything Model (SAM) for radiotherapy OAR segmentation.
  • To assess SAM's generalizability across multiple disease sites and its effectiveness with interactive prompting.

Summary:

  • SAM's 'segment anything' mode achieved clinically acceptable OAR segmentation (Dice > 0.7) across prostate, lung, gastrointestinal, and head & neck sites.
  • The 'box prompt' mode significantly improved segmentation accuracy (Dice +0.1–0.5) through interactive, on-the-fly adjustments.
Keywords:
artificial intelligenceclinical delineationgeneralizablepromptableradiation oncologysegment anything model

Related Experiment Videos

  • Performance varied by organ size and boundary distinctiveness, with better results for large, well-defined organs like lungs and poorer results for smaller, less distinct organs like parotid.

    • Impact:

    • Demonstrates SAM's robust generalizability and consistent accuracy for automatic radiotherapy segmentation.
    • The interactive box-prompt method facilitates dynamic, patient-specific auto-segmentation, enhancing treatment planning efficiency.
    • SAM's cross-site and cross-modality capabilities support the development of a universal auto-segmentation model for radiotherapy.