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

Spectral clustering algorithms for ultrasound image segmentation.

Neculai Archip1, Robert Rohling, Peter Cooperberg

  • 1Computational Radiology Laboratory, Harvard Medical School, Department of Radiology, Brigham and Women's Hospital, Boston, MA, USA. narchip@bwh.harvard.edu

Medical Image Computing and Computer-Assisted Intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention
|May 12, 2006
PubMed
Summary

This study introduces a novel unsupervised image segmentation method based on spectral clustering for ultrasound images. The new technique shows promise for segmenting abdominal and fetal medical images.

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

An MRI Atlas of the Human Fetal Brain: Reference and Segmentation Tools for Fetal Brain MRI Analysis.

Scientific data·2026
Same author

Deep transfer learning for detection of placenta-mediated diseases from ultrasound images.

Scientific reports·2026
Same author

Impact of Synthetic Lesional MR Images in Automated Focal Cortical Dysplasia Detection in Low-Data Scenarios.

Journal of neuroimaging : official journal of the American Society of Neuroimaging·2026
Same author

A two-step temporal data augmentation and supervised learning framework for predicting autism diagnosis at 36 months in patients with tuberous sclerosis complex.

Computers in biology and medicine·2026
Same author

Comparison of Myeloarchitectonic Feature Recognition of the Primary Visual Cortex at 7 T Relative to 3 T MRI.

Journal of magnetic resonance imaging : JMRI·2026
Same author

Super-resolution MRI-derived brainstem and cerebellar volumes in fetuses between 22 weeks and 32 weeks of gestation.

Pediatric radiology·2026

Area of Science:

  • Medical imaging
  • Computer vision
  • Signal processing

Background:

  • Spectral clustering algorithms utilize image graph Laplacian eigenvectors for segmentation.
  • Normalized Cut (NCut) is a supervised method for image segmentation.
  • Unsupervised segmentation of medical images, particularly ultrasound, remains a challenge.

Purpose of the Study:

  • To investigate the suitability of unsupervised spectral clustering for ultrasound image segmentation.
  • To develop and evaluate a novel unsupervised segmentation algorithm extending the NCut criterion.
  • To compare the novel algorithm's performance against manual segmentation and the classical NCut method.

Main Methods:

  • Extension of the Normalized Cut (NCut) technique to unsupervised clustering.

Related Experiment Videos

  • Application of the novel segmentation algorithm to simulated ultrasound images.
  • Validation on abdominal and fetal ultrasound images, with comparisons to manual segmentation and classical NCut.

    • Main Results:

    • The developed unsupervised segmentation technique was applied to simulated and real ultrasound images.
    • Segmentation results were compared with manual segmentations and the traditional NCut algorithm.
    • Performance was evaluated on abdominal and fetal ultrasound datasets, and other medical image types.

    Conclusions:

    • The proposed unsupervised spectral clustering approach demonstrates potential for segmenting ultrasound images.
    • The method shows comparable or improved results to manual segmentation and classical NCut in initial tests.
    • Further research is warranted to fully establish its clinical utility across diverse medical imaging applications.