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 Concept Videos

You might also read

Related Articles

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

Sort by
Same author

Comparison of Measurement Techniques for Photoreceptor Loss in Geographic Atrophy.

Translational vision science & technology·2026
Same author

Neuroretinal Layer Thinning on OCT Imaging and Hemoglobin A1c in Youth With Type 1 Diabetes.

JAMA ophthalmology·2026
Same author

Comparison of Clarus, Optos, and Heidelberg Systems for Geographic Atrophy Area Measurements.

Ophthalmology science·2026
Same author

Reference Standard for Validation of Age-Related Macular Degeneration Screening Algorithms.

Ophthalmology·2026
Same author

Modest and Variable Correlations Between Geographic Atrophy Enlargement Rates in Fellow Eyes in the AREDS2 Study.

JAMA ophthalmology·2026
Same author

Geographic Atrophy on Fundus Autofluorescence and Color Fundus Photographs: Association with Deep Visual Sensitivity Losses.

Ophthalmology science·2026

Related Experiment Video

Updated: Aug 16, 2025

Deep Learning-Based Segmentation of Cryo-Electron Tomograms
10:25

Deep Learning-Based Segmentation of Cryo-Electron Tomograms

Published on: November 11, 2022

9.1K

Implementation of a Large-Scale Image Curation Workflow Using Deep Learning Framework.

Amitha Domalpally1,2, Robert Slater1, Nancy Barrett2

  • 1A-EYE Unit, Department of Ophthalmology and Visual Sciences, University of Wisconsin, Madison, Wisconsin.

Ophthalmology Science
|December 19, 2022
PubMed
Summary
This summary is machine-generated.

This study developed an artificial intelligence (AI) algorithm for efficient image curation, achieving 88.3% agreement with human graders. The AI system successfully flags potential mislabels, significantly reducing error rates in high-volume settings.

Keywords:
AI, artificial intelligenceArtificial intelligenceDeep learningFundus photographImage curationMachine learningMetadataRetinal imageStandardization

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

9.3K
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

2.9K

Related Experiment Videos

Last Updated: Aug 16, 2025

Deep Learning-Based Segmentation of Cryo-Electron Tomograms
10:25

Deep Learning-Based Segmentation of Cryo-Electron Tomograms

Published on: November 11, 2022

9.1K
Deep Neural Networks for Image-Based Dietary Assessment
13:19

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

9.3K
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

2.9K

Area of Science:

  • Ophthalmology
  • Artificial Intelligence
  • Medical Imaging

Background:

  • Human-powered image curation is essential but time-consuming for developing artificial intelligence (AI) algorithms.
  • High-volume clinical settings require efficient and accurate image curation methods.

Purpose of the Study:

  • To develop and implement an AI algorithm for image curation in a high-volume setting.
  • To explore AI tools for a tiered image review approach, flagging potential mislabels for human verification.

Main Methods:

  • An AI algorithm was developed to classify seven-field stereoscopic retinal images into 8 field numbers.
  • Probability scores were generated to identify potentially mislabeled images, with a cutoff score determined using receiver operating characteristic curves.

Main Results:

  • The AI model achieved 88.3% agreement (kappa, 0.87) with human grader classifications on a test set of 3004 images.
  • A probability score cutoff of 0.99 effectively distinguished mislabeled images, reducing the error rate from 11.7% to 1.5% in a tiered review workflow.

Conclusions:

  • AI algorithms require additional measures beyond validation, such as error-flagging tools, to enhance accuracy and trust.
  • Implementing AI-driven image curation with human review significantly improves efficiency and reduces errors in large-scale datasets.