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 Video

Updated: Jun 9, 2025

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

370

Plug-and-play segment anything model improves nnUNet performance.

Yunxiang Li1, Bowen Jing1, Zihan Li2

  • 1Department of Radiation Oncology, UT Southwestern Medical Center, Dallas, Texas, USA.

Medical Physics
|October 28, 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

Promera: a unified model for biomolecular structure prediction, filtering, and design.

bioRxiv : the preprint server for biology·2026
Same author

Thousandfold Expansion Microscopy.

bioRxiv : the preprint server for biology·2026
Same author

SwitchCraft: A Programmatic Framework for Designing State-Switching Proteins.

ArXiv·2026
Same author

AI-based methods for simulating, sampling, and predicting protein ensembles.

Current opinion in structural biology·2026
Same author

Learning the language of protein-protein interactions.

Nature communications·2026
Same author

Improvement in breast lesion classification utilizing deep learning and treatment response assessment maps (TRAMs).

Proceedings of SPIE--the International Society for Optical Engineering·2025

nnSAM enhances medical image segmentation by combining Segment Anything Model (SAM) and nnUNet for automated, accurate results on small datasets. This approach significantly improves segmentation performance, especially with limited training data.

Area of Science:

  • Medical Imaging
  • Computer Vision
  • Machine Learning

Background:

  • Automatic medical image segmentation is crucial for clinical workflows.
  • Foundational models like Segment Anything Model (SAM) offer universal segmentation but require human interaction and domain adaptation.
  • Traditional models (e.g., nnUNet) automate segmentation but need extensive domain-specific training data.

Purpose of the Study:

  • To develop a fully automated segmentation method for limited training samples by integrating SAM and nnUNet.
  • To enhance accuracy and robustness in medical image segmentation using small datasets.
  • To leverage the strengths of both foundational and domain-specific models.

Main Methods:

  • Proposed nnSAM model for small sample medical image segmentation.
Keywords:
few‐shot learningfoundation modelmedical image segmentation

More Related Videos

Deep Neural Networks for Image-Based Dietary Assessment
13:19

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

9.0K
Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application
05:56

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Published on: April 14, 2023

2.4K

Related Experiment Videos

Last Updated: Jun 9, 2025

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

370
Deep Neural Networks for Image-Based Dietary Assessment
13:19

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

9.0K
Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application
05:56

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Published on: April 14, 2023

2.4K
  • Integrated SAM's feature extraction with nnUNet's automatic configuration for robust adaptation on small datasets.
  • Implemented a boundary shape supervision loss using level set functions and curvature calculations to learn anatomical priors from limited annotations.

    • Main Results:

    • nnSAM demonstrated superior performance across brain white matter, liver, lung, and heart segmentation tasks.
    • In brain white matter segmentation with 20 samples, nnSAM achieved a DICE score of 82.77% and ASD of 1.14 mm, outperforming nnUNet.
    • The performance advantage of nnSAM was more pronounced with fewer training samples.

    Conclusions:

    • nnSAM significantly improves segmentation performance on small-sample tasks.
    • The study highlights the potential of small-sample learning in medical image segmentation.
    • nnSAM offers a promising solution for automated medical image segmentation with limited data.