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: Jun 16, 2026

Using Retinal Imaging to Study Dementia
09:17

Using Retinal Imaging to Study Dementia

Published on: November 6, 2017

From discarding to leveraging: quality-aware collaborative learning for robust diabetic retinopathy grading.

Yuan Pan1, Xiangwen Cai2, Pan Xiong1

  • 1School of Medical Information Engineering, Guangzhou University of Chinese Medicine, Guangzhou, China.

Frontiers in Medicine
|June 15, 2026
PubMed
Summary

Related Concept Videos

Diabetic Retinopathy01:27

Diabetic Retinopathy

DefinitionDiabetic retinopathy is a microvascular complication of diabetes affecting the retinal blood vessels.Risk FactorsDiabetic retinopathy is present in almost all individuals with type 1 diabetes and more than 60% of those with type 2 diabetes after two decades of disease.The risk increases with poor glycemic control, hypertension, dyslipidemia, smoking, pregnancy, and puberty.Although cataracts and glaucoma are also more frequent in people with diabetes, retinopathy remains the leading...

You might also read

Related Articles

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

Sort by
Same author

Bio-Inspired Dual-Gradient Aerogel Enables Sustainable Freshwater Production and Energy Harvesting.

Advanced materials (Deerfield Beach, Fla.)·2026
Same author

Optimizing frozen section sampling for thyroid follicular neoplasms through intraoperative ultrasound guidance.

European thyroid journal·2026
Same author

Allergic Sensitization to Common Ragweed and Mugwort Pollen Allergens: A 6-Year Single-Center Retrospective Analysis.

Journal of immunology research·2026
Same author

Unraveling the Design Principles of Crystallographic Orientation for Ultralong, 8500 Cycled Fe-Ion Battery.

Advanced materials (Deerfield Beach, Fla.)·2026
Same author

A highly stretchable tri-channel fiber for composite motion decoupling.

Nature communications·2026
Same author

Bridging research and practice: a systematic review of Academic entrepreneurship in healthcare and its emerging research agenda.

BMC health services research·2026

Diabetic retinopathy screening struggles with low-quality images showing severe disease. A new quality-guided framework (QGDR) uses image quality as a continuous signal, improving diagnostic accuracy for all image types.

Area of Science:

  • Ophthalmology
  • Medical Imaging
  • Artificial Intelligence

Background:

  • Diabetic retinopathy (DR) screening is challenged by poor image quality in severe cases due to disease-related opacities.
  • Conventional methods often discard critical low-quality images or misclassify them, hindering effective screening.
  • Image quality assessment (IQA) is typically a binary preprocessing step, failing to leverage quality information dynamically.

Purpose of the Study:

  • To develop and validate a novel framework, QGDR, that utilizes image quality as a continuous guidance signal for DR diagnosis.
  • To address the paradox of low-quality images often containing the most severe DR indicators.
  • To improve the reliability and coverage of automated DR screening systems.

Main Methods:

  • Proposed QGDR, a quality-guided dynamic routing framework incorporating multi-level IQA, quality-conditioned context gating, and adaptive gated fusion.
Keywords:
collaborative learningdiabetic retinopathy gradingdynamic expert routingfundus image quality assessmentmulti-scale feature learning

Related Experiment Videos

Last Updated: Jun 16, 2026

Using Retinal Imaging to Study Dementia
09:17

Using Retinal Imaging to Study Dementia

Published on: November 6, 2017

  • Developed a multi-level IQA module to extract hierarchical quality features.
  • Implemented a dynamic routing mechanism that routes inputs to scale-specialized experts based on predicted image quality.

    • Main Results:

    • QGDR achieved high accuracy (78.32% on EyeQ, 80.85% on DDR) and outperformed baseline models including CNNs, transformers, and foundation models.
    • Performance was maintained or improved on low-quality images (e.g., 82.21% on DDR-Reject), exceeding good-quality accuracy.
    • Validated cross-dataset generalization on IDRiD and DeepDRiD cohorts, confirming robustness.

    Conclusions:

    • Treating image quality as a continuous guidance signal, rather than a filter, enhances DR diagnostic reliability.
    • The QGDR framework effectively leverages semantic quality information to improve screening for both high- and low-quality images.
    • QGDR offers a promising solution for preserving screening coverage without compromising diagnostic accuracy in real-world DR screening.