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: Jul 9, 2026

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

SEGMENTATION CONFIDENCE FOR ARBITRARY CNNS.

Baris Oguz1, Xing Yao2, Claudia Tawil1

  • 1University of Pennsylvania.

Proceedings. IEEE International Symposium on Biomedical Imaging
|July 8, 2026
PubMed
Summary

Convolutional neural networks (CNNs) for medical image segmentation lack confidence measures. This study introduces a novel method for voxel-based uncertainty estimation and a confidence score that predicts segmentation performance on new data without retraining.

Related Concept Videos

Confidence Coefficient01:24

Confidence Coefficient

The confidence coefficient is also known as the confidence level or degree of confidence. It is the percent expression for the probability, 1-α, that the confidence interval contains the true population parameter assuming that the confidence interval is obtained after sufficient unbiased sampling; for example, if the CL = 90%, then in 90 out of 100 samples the interval estimate will enclose the true population parameter. Here α is the area under the curve, distributed equally under both the...

You might also read

Related Articles

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

Sort by
Same author

LOTUS: Latent Outpainting Diffusion Model for Three-Dimensional Ultrasound Stitching.

Proceedings of machine learning research·2026
Same author

VFMStitch: A Vision-Foundation-Model Empowered Framework for 3D Ultrasound Stitching via Geometric-Semantic Feature Fusion.

Proceedings of machine learning research·2026
Same author

From Geometry to Intensity: A Coarse-to-Fine Pipeline for Unsupervised 3D Ultrasound Stitching.

Proceedings of SPIE--the International Society for Optical Engineering·2026
Same author

Endo-SemiS: Towards Robust Semi-Supervised Image Segmentation for Endoscopic Video.

Proceedings of machine learning research·2026
Same author

Kidney Endoscopy Video to Preoperative CT Alignment for Depth Estimation.

Proceedings of SPIE--the International Society for Optical Engineering·2026
Same author

AVA: Automated Viewability Analysis for Ureteroscopic Intrarenal Surgery.

Proceedings of SPIE--the International Society for Optical Engineering·2026

Area of Science:

  • Medical image analysis
  • Machine learning in healthcare
  • Deep learning for segmentation

Background:

  • Convolutional neural networks (CNNs) are essential for medical image segmentation.
  • Current CNNs lack inherent confidence estimation, hindering clinical trust.
  • Existing uncertainty methods require custom architectures and retraining.

Purpose of the Study:

  • To develop a method for voxel-based uncertainty estimation in CNN segmentation.
  • To create a straightforward confidence score from uncertainty maps.
  • To enable uncertainty estimation without retraining existing CNN models.

Main Methods:

  • Exploration of voxel-based uncertainty estimation techniques.
  • Development of a novel method to derive a single confidence score from uncertainty maps.
Keywords:
Confidencesegmentationuncertainty

Related Experiment Videos

Last Updated: Jul 9, 2026

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

  • Evaluation across diverse regions of interest (ROIs) and medical imaging modalities.

    • Main Results:

    • Proposed methods can be integrated into existing segmentation pipelines without retraining.
    • The novel confidence score effectively predicts segmentation performance on unseen data.
    • Demonstrated effectiveness across different ROIs and modalities.

    Conclusions:

    • The developed approach provides a practical way to add confidence estimation to CNN-based medical image segmentation.
    • The proposed confidence score is a reliable indicator of segmentation quality for new datasets.
    • This work enhances the reliability and interpretability of deep learning models in medical imaging.