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

Characterizing Response to PARP Inhibitor Treatment Combinations in Advanced Prostate Cancer.

Biomedicines·2026
Same author

Predictors of Good Visual Recovery in Patients With Spontaneously Resolved Acute CSCR - MICRoN Report Number Thirteen.

American journal of ophthalmology·2026
Same author

Real-world durability challenges with nAMD treatments and the potential promise of gene therapy.

Current medical research and opinion·2026
Same author

Congenital Optic Nerve Anomalies and Associated Systemic Conditions.

International ophthalmology clinics·2026
Same author

Factors Associated with Machine Learning-Based Predictions of Retinal Aging Using Teleretinal Screening Images from Patients with Diabetes.

Ophthalmology science·2026
Same author

Simultaneous Segmentation of Geographic Atrophy in Longitudinally Acquired Fundus Autofluorescence Images.

Ophthalmology science·2026

Related Experiment Video

Updated: Sep 15, 2025

Author Spotlight: Ex Vivo OCT-Based Multimodal Imaging of Human Donor Eyes for Research into Age-Related Macular Degeneration
10:14

Author Spotlight: Ex Vivo OCT-Based Multimodal Imaging of Human Donor Eyes for Research into Age-Related Macular Degeneration

Published on: May 26, 2023

3.6K

Genotype Prediction from Retinal Fundus Images Using Deep Learning in Eyes with Age-Related Macular Degeneration.

Avishai Halev1, Denis Huang2, Shahbaz Rezaei3

  • 1Department of Mathematics, University of California, Davis, Davis, California.

Ophthalmology Science
|July 15, 2025
PubMed
Summary

Deep learning models can predict high-risk genetic variants for age-related macular degeneration (AMD) from retinal images. This noninvasive approach offers insights into genotype-phenotype relationships in AMD.

Keywords:
AMDAge-related macular degenerationDeep learningGenotype predictionMachine learning

More Related Videos

Author Spotlight: Unraveling the Pathogenesis of Age-Related Macular Degeneration and Discovering Potential Therapies
06:16

Author Spotlight: Unraveling the Pathogenesis of Age-Related Macular Degeneration and Discovering Potential Therapies

Published on: July 28, 2023

2.7K
Author Spotlight: Understanding Age-Related Macular Degeneration Pathophysiology with QAF Workflow
08:54

Author Spotlight: Understanding Age-Related Macular Degeneration Pathophysiology with QAF Workflow

Published on: May 26, 2023

1.7K

Related Experiment Videos

Last Updated: Sep 15, 2025

Author Spotlight: Ex Vivo OCT-Based Multimodal Imaging of Human Donor Eyes for Research into Age-Related Macular Degeneration
10:14

Author Spotlight: Ex Vivo OCT-Based Multimodal Imaging of Human Donor Eyes for Research into Age-Related Macular Degeneration

Published on: May 26, 2023

3.6K
Author Spotlight: Unraveling the Pathogenesis of Age-Related Macular Degeneration and Discovering Potential Therapies
06:16

Author Spotlight: Unraveling the Pathogenesis of Age-Related Macular Degeneration and Discovering Potential Therapies

Published on: July 28, 2023

2.7K
Author Spotlight: Understanding Age-Related Macular Degeneration Pathophysiology with QAF Workflow
08:54

Author Spotlight: Understanding Age-Related Macular Degeneration Pathophysiology with QAF Workflow

Published on: May 26, 2023

1.7K

Area of Science:

  • Ophthalmology
  • Genetics
  • Artificial Intelligence

Background:

  • Age-related macular degeneration (AMD) is a leading cause of vision loss.
  • Genetic factors, particularly in the complement factor H (CFH) and age-related maculopathy susceptibility 2 (ARMS2) genes, significantly influence AMD risk.
  • Current methods for genetic risk assessment can be invasive or require specialized laboratory analysis.

Purpose of the Study:

  • To develop and evaluate deep learning models for predicting high-risk genetic variants associated with AMD directly from retinal fundus photographs.
  • To classify single-nucleotide polymorphisms (SNPs) in the CFH and ARMS2 genes using retinal images.

Main Methods:

  • Utilized a dataset of 31,271 retinal color fundus photographs from 1,754 participants in the Age-Related Eye Disease Study.
  • Trained deep learning models, including Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs), to classify patients into high-risk or low-risk genotypes for CFH and ARMS2.
  • Evaluated model performance on an independent test set and used attribution mapping to identify relevant image features.

Main Results:

  • Vision Transformer (ViT) models achieved Area Under the Receiver Operating Characteristic Curve (AUROC) of 0.719 for CFH and 0.741 for ARMS2 high-risk genotype prediction.
  • Model performance for ARMS2 genotype prediction was significantly higher in eyes with advanced AMD (AUROC 0.867), choroidal neovascularization (AUROC 0.833), and geographic atrophy (AUROC 0.957).
  • Genotype prediction from fundus images was more challenging than AMD severity or gender classification, but saliency mapping indicated attention to the macula.

Conclusions:

  • Deep learning models can effectively predict high-risk genotypes in CFH and ARMS2 from noninvasive retinal fundus images.
  • These findings demonstrate the potential for inferring genetic predisposition to AMD from ocular imaging.
  • The study provides valuable insights into genotype-phenotype correlations in AMD, paving the way for novel diagnostic and risk assessment strategies.