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

Simultaneous truth and performance level estimation (STAPLE): an algorithm for the validation of image segmentation.

Simon K Warfield1, Kelly H Zou, William M Wells

  • 1Harvard Medical School and the Department of Radiology of Brigham and Women's Hospital, 75 Francis St, Boston, MA 02115, USA. warfield@bwh.harvard.edu

IEEE Transactions on Medical Imaging
|July 15, 2004
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

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

Scientific data·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
Same author

LesionSCynth: A simple parametric lesion synthesis method to improve spinal cord lesion segmentation in low-data scenarios.

Imaging neuroscience (Cambridge, Mass.)·2025
Same journal

MUST: Multi-style virtual staining with incomplete pairs.

IEEE transactions on medical imaging·2026
Same journal

BrainCL: Transformer-Based Brain Network Contrastive Learning with Multi-Order Topology and Salience Masking.

IEEE transactions on medical imaging·2026
Same journal

LLM-enhanced Neuron Segmentation and Reconstruction in Complex Mouse Brain Images.

IEEE transactions on medical imaging·2026
Same journal

Matrixed-Spectrum Decomposition Accelerated Linear Boltzmann Transport Equation Solver for Fast Scatter Correction in Multi-Spectral CT.

IEEE transactions on medical imaging·2026
Same journal

The Ritz Adjoint Method for MRI Pulse Design.

IEEE transactions on medical imaging·2026
Same journal

Physiology-guided Self-supervised Learning for Simultaneous Dual-Tracer PET Separation.

IEEE transactions on medical imaging·2026
See all related articles

Evaluating image segmentation performance is challenging due to rater variability and lack of ground truth. The Simultaneous Truth and Performance Level Estimation (STAPLE) algorithm addresses this by estimating true segmentations and rater performance from multiple inputs.

Area of Science:

  • Medical Image Analysis
  • Computational Pathology
  • Biomedical Engineering

Background:

  • Assessing medical image segmentation performance is difficult due to limited accuracy and precision of algorithms.
  • Human rater variability (intra- and inter-rater) complicates performance evaluation, despite being the gold standard.
  • Lack of ground truth in clinical data hinders objective quantification of segmentation accuracy.

Purpose of the Study:

  • To develop a robust method for evaluating image segmentation performance on clinical data.
  • To simultaneously estimate the ground truth segmentation and the performance level of individual raters (human or algorithmic).
  • To enable direct comparison between human and automated segmentation algorithm performance.

Main Methods:

  • An expectation-maximization algorithm, Simultaneous Truth and Performance Level Estimation (STAPLE), was developed.

Related Experiment Videos

  • STAPLE computes a probabilistic estimate of the true segmentation by optimally combining multiple segmentations.
  • Performance levels of individual raters are estimated and used to weight their contributions to the final segmentation estimate.

    • Main Results:

    • STAPLE provides a probabilistic estimate of the true segmentation, incorporating prior spatial models and homogeneity constraints.
    • The algorithm quantifies the performance level of each contributing segmentation (rater or algorithm).
    • It facilitates direct assessment of automated image segmentation algorithms and comparison with human raters on clinical data.

    Conclusions:

    • STAPLE offers a practical solution for evaluating image segmentation performance directly on clinical data.
    • The algorithm effectively addresses the challenge of ground truth estimation in medical imaging.
    • STAPLE enables objective assessment and comparison of both human and automated segmentation methods.