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

Role of DOTATATE PET Imaging in Meningiomas with Neuro-Ophthalmic Implications.

Neuro-ophthalmology (Aeolus Press)·2026
Same author

Retinal lesion annotation on fundus imaging: an interobserver variability study.

Scientific reports·2026
Same author

Cell type transcriptomic modules reveal shared molecular mechanisms in Alzheimer's and Parkinson's disease.

GigaScience·2026
Same author

DNAi: an open-source AI tool for unbiased DNA fiber analysis.

Nucleic acids research·2026
Same author

Intraoperative Optical Coherence Tomography Features of Epiretinal Human Amniotic Membrane Graft Under Different Tamponade Agents.

Journal of vitreoretinal diseases·2026
Same author

DTG: Dual transformers-based generative adversarial networks for retinal 2D/3D OCT image classification.

Medical image analysis·2026

Related Experiment Video

Updated: Jun 27, 2025

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

2.7K

Comparing code-free deep learning models to expert-designed models for detecting retinal diseases from optical

Samir Touma1,2,3, Badr Ait Hammou1,2, Fares Antaki1,2,3,4

  • 1Department of Ophthalmology, Université de Montréal, Montreal, Québec, Canada.

International Journal of Retina and Vitreous
|April 26, 2024
PubMed
Summary

Code-free deep learning (CFDL) models developed by ophthalmologists achieved performance comparable to expert-designed AI models in detecting retinal diseases from OCT scans. This demonstrates CFDL

Keywords:
Artificial intelligenceCode-free machine learningOptical coherence tomography

More Related Videos

Application of Optical Coherence Tomography to a Mouse Model of Retinopathy
08:22

Application of Optical Coherence Tomography to a Mouse Model of Retinopathy

Published on: January 12, 2022

4.0K
In vivo Structural Assessments of Ocular Disease in Rodent Models using Optical Coherence Tomography
07:44

In vivo Structural Assessments of Ocular Disease in Rodent Models using Optical Coherence Tomography

Published on: July 24, 2020

2.9K

Related Experiment Videos

Last Updated: Jun 27, 2025

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

2.7K
Application of Optical Coherence Tomography to a Mouse Model of Retinopathy
08:22

Application of Optical Coherence Tomography to a Mouse Model of Retinopathy

Published on: January 12, 2022

4.0K
In vivo Structural Assessments of Ocular Disease in Rodent Models using Optical Coherence Tomography
07:44

In vivo Structural Assessments of Ocular Disease in Rodent Models using Optical Coherence Tomography

Published on: July 24, 2020

2.9K

Area of Science:

  • Ophthalmology
  • Artificial Intelligence
  • Medical Imaging

Background:

  • Code-free deep learning (CFDL) offers a novel approach in artificial intelligence (AI).
  • This study evaluates CFDL models built by clinicians versus expert-developed AI models for retinal pathology detection.
  • Optical coherence tomography (OCT) videos and images are used for analysis.

Purpose of the Study:

  • To compare the diagnostic performance of CFDL models created by ophthalmologists without coding experience against bespoke AI models.
  • To assess the accuracy of CFDL in classifying retinal pathologies from OCT data.

Main Methods:

  • A dataset of 1,173 OCT macular videos and fovea-centered images was used.
  • Models were developed independently by an ophthalmology resident (CFDL) and an AI expert (bespoke).
  • A multi-class model was designed to identify five retinal conditions: normal, macular hole, epiretinal membrane, wet age-related macular degeneration, and diabetic macular edema.

Main Results:

  • CFDL models showed excellent discriminative performance, with some metrics surpassing bespoke models for video analysis.
  • The fovea-centered CFDL model outperformed the video-based model and matched the accuracy of the best bespoke model.
  • Specific performance metrics included area under the precision-recall curve (0.984 for CFDL video vs. 0.901 for bespoke video), precision (94.1% vs. 94.2%), sensitivity (94.1% vs. 94.2%), and accuracy (94.1% vs. 96.7%).

Conclusions:

  • Clinician-developed CFDL models achieve performance on par with expert-designed bespoke models for retinal pathology classification.
  • CFDL shows promise for democratizing AI in medicine, enabling wider clinical adoption.
  • Addressing the limitations of CFDL is crucial for its effective integration into healthcare settings.