Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

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

Plain language summary of publication of the 96-week results from the PHOTON study: intravitreal aflibercept 8 mg for diabetic macular edema.

Therapeutic advances in ophthalmology·2026

Same author

Nanoparticle-mediated inhibition of Yes-associated protein prevents corneal scarring after traumatic injury.

Materials today. Bio·2026

Same author

<i>In situ</i>rheological monitoring of diffusion-controlled hydrogel crosslinking for embedded 3D bioprinting.

Biofabrication·2026

Same author

Intraocular Pressure Outcomes with Intravitreal Aflibercept 8 mg: A 96-Week Pooled Analysis from PULSAR and PHOTON.

Ophthalmology and therapy·2026

Same author

Safety and Biodistribution of Topical Laquinimod, an Immunomodulator Targeting Aryl Hydrocarbon Receptor: The LION Study.

Ophthalmology science·2026

Same author

Response to Letter Regarding "Long-Term Outcomes of Corneal Crosslinking for Keratoconus Across Pediatric, Adolescent, and Adult Populations in a Diverse U.S. Population".

Journal of cataract and refractive surgery·2026

Search research articles

Related Experiment Video

Updated: Sep 3, 2025

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma

Published on: October 10, 2018

Predicting Systemic Health Features from Retinal Fundus Images Using Transfer-Learning-Based Artificial Intelligence

Nergis C Khan¹, Chandrashan Perera^1,2, Eliot R Dow¹

¹Byers Eye Institute at Stanford, Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA 94305, USA.

Diagnostics (Basel, Switzerland)

|July 27, 2022

Summary

Ocular imaging with deep learning (DL) artificial intelligence (AI) can predict systemic health features. Transfer learning from general image datasets like ImageNet significantly improved AI model accuracy for predicting features from fundus images.

Keywords:

artificial intelligence diabetic retinopathy retinal imaging transfer learning

More Related Videos

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Published on: October 27, 2023

Using Retinal Imaging to Study Dementia

Using Retinal Imaging to Study Dementia

Published on: November 6, 2017

Related Experiment Videos

Last Updated: Sep 3, 2025

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma

Published on: October 10, 2018

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Published on: October 27, 2023

Using Retinal Imaging to Study Dementia

Using Retinal Imaging to Study Dementia

Published on: November 6, 2017

Area of Science:

Ophthalmology
Medical Imaging
Artificial Intelligence

Background:

Color fundus photography is standard for diagnosing eye conditions.
Ocular imaging data may reveal systemic health insights.
Computer vision and DL/AI models can analyze these insights.

Purpose of the Study:

To develop a DL model for predicting systemic features from fundus images.
To determine the optimal method for constructing such DL models.
To assess the predictive performance of different transfer learning approaches.

Main Methods:

Collected data from diabetic retinopathy screening patients (March 2020-March 2021).
Developed DenseNet201 DL models using two transfer learning strategies: ImageNet and fundus images.
Trained models on 1277 fundus images and compared performance using Area Under the Receiver Operating Characteristics Curve (AUROC).

Main Results:

ImageNet transfer learning models outperformed fundus image transfer learning models (mean AUROC 0.78 vs. 0.65, p < 0.001).
ImageNet pretraining enabled prediction of ethnicity (AUROC 0.93), age > 70 (AUROC 0.90), gender (AUROC 0.85), ACE inhibitor use (AUROC 0.82), and ARB medication use (AUROC 0.78).

Conclusions:

Fundus images contain significant information about a patient's systemic health.
Transfer learning from generalized image datasets (like ImageNet) is recommended to enhance DL model accuracy for medical image analysis.
This approach optimizes DL model performance for predicting systemic features from ocular imaging.