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

Statistical validation of image segmentation quality based on a spatial overlap index.

Kelly H Zou1, Simon K Warfield, Aditya Bharatha

  • 1Department of Radiology and Surgical Planning Laboratory, Brigham and Women's Hospital, Harvard Medical School, 75 Francis St (Floor L-l), Boston, MA 02115, USA.

Academic Radiology
|February 21, 2004
PubMed
Summary

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

Missed radiologic disease activity by conventional side-by-side MRI comparison is associated with future disease activity in multiple sclerosis: Adopting semiautomated coregistration-fusion into clinical practice.

Multiple sclerosis (Houndmills, Basingstoke, England)·2026
Same author

Plasma CHI3L1 as a marker of cognitive and structural brain changes in radiologically isolated syndrome.

BMJ neurology open·2026
Same author

Conventional and Emerging MRI Biomarkers of Multiple Sclerosis.

Radiographics : a review publication of the Radiological Society of North America, Inc·2026
Same author

AI-assisted cryo-dose estimation and tissue injury modeling for assessing side effects in MRI-guided prostate cryoablation: a retrospective study.

Physics in medicine and biology·2026
Same author

Paramagnetic Rim Lesions and Development of Clinical MS in Radiologically Isolated Syndrome.

JAMA neurology·2026
Same author

Choroid plexus enlargement correlates with demyelinating lesions and thalamic atrophy in radiologically isolated syndrome.

Multiple sclerosis (Houndmills, Basingstoke, England)·2025

The Dice Similarity Coefficient (DSC) effectively measures spatial overlap for validating image segmentation accuracy and reproducibility in medical imaging. Results showed generally good but variable performance in prostate and brain tumor MR imaging applications.

Area of Science:

  • Medical Imaging Analysis
  • Biomedical Engineering
  • Radiology

Background:

  • Accurate segmentation of anatomical structures in medical images is crucial for diagnosis and treatment planning.
  • Evaluating the performance of segmentation methods, both manual and automated, requires robust statistical validation metrics.
  • The Dice Similarity Coefficient (DSC) is a widely used metric for assessing overlap between segmentations.

Purpose of the Study:

  • To evaluate the Dice Similarity Coefficient (DSC) as a statistical method for validating image segmentation.
  • To assess the reproducibility of manual segmentations and the accuracy of automated segmentations using DSC.
  • To demonstrate the application of DSC in clinical scenarios involving prostate and brain tumor MR imaging.

Main Methods:

Related Experiment Videos

  • The Dice Similarity Coefficient (DSC) was employed to quantify spatial overlap.
  • Manual segmentations of the prostate peripheral zone were repeated on preoperative and intraoperative MRI scans.
  • A semi-automated probabilistic fractional segmentation algorithm was applied to MR images of brain tumors.
  • Statistical analysis included the analysis of variance (ANOVA) for logit-transformed DSC values.

    • Main Results:

    • Prostate segmentation reproducibility showed good performance (mean DSC 0.883 for 1.5T, 0.838 for 0.5T MRI).
    • Brain tumor segmentation yielded a wide range of DSC values across different tumor types (Meningiomas, astrocytomas, gliomas).
    • Significant differences in DSC were observed between different imaging conditions and tumor types.

    Conclusions:

    • The DSC is a valuable and straightforward metric for assessing spatial overlap in image segmentation studies.
    • DSC demonstrated generally satisfactory but variable results in validating manual and automated segmentation tasks.
    • The DSC metric is adaptable for various validation tasks in medical image analysis.