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

Association between myosteatosis and bone mineral density in postmenopausal women.

Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA·2026
Same author

Refractive and Visual Outcomes of SmartSight NOVA Refractive Lenticule Extraction for Myopia and Myopic Astigmatism at 3 Months: a Retrospective Case Series.

Ophthalmology and therapy·2026
Same author

Oculomics of lipid metabolism: A scoping review across anterior and posterior segment diseases.

Science progress·2026
Same author

Structural variation of the pineal gland in shift workers from neuroimaging evidence.

Chronobiology international·2026
Same author

Size exchange and vault response in posterior chamber phakic intraocular lens implantation: 12 years at a high-volume center.

Journal of cataract and refractive surgery·2026
Same author

Clinical use and radiological yield of magnetic resonance fingerprinting in epilepsy.

Epilepsia·2026

Related Experiment Video

Updated: Dec 22, 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

3.2K

CycleGAN-based deep learning technique for artifact reduction in fundus photography.

Tae Keun Yoo1, Joon Yul Choi2, Hong Kyu Kim3

  • 1Department of Ophthalmology, Medical Research Center, Aerospace Medical Center, Republic of Korea Air Force, 635 Danjae-ro, Sangdang-gu, Cheongju, South Korea. eyetaekeunyoo@gmail.com.

Graefe'S Archive for Clinical and Experimental Ophthalmology = Albrecht Von Graefes Archiv Fur Klinische Und Experimentelle Ophthalmologie
|May 4, 2020
PubMed
Summary

This study demonstrates how a CycleGAN deep learning model effectively removes artifacts from fundus photographs. The technique improves image quality, aiding clinicians in diagnosing retinal conditions from low-quality images.

Keywords:
ArtifactDeep learningFundus photographyGenerative adversarial networkImage quality

More Related Videos

Quantitative Fundus Autofluorescence for the Evaluation of Retinal Diseases
07:22

Quantitative Fundus Autofluorescence for the Evaluation of Retinal Diseases

Published on: March 11, 2016

11.8K
Retrospective Cardiac Gating with A Prototype Small-Animal X-ray Computed Tomograph
05:32

Retrospective Cardiac Gating with A Prototype Small-Animal X-ray Computed Tomograph

Published on: February 21, 2025

611

Related Experiment Videos

Last Updated: Dec 22, 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

3.2K
Quantitative Fundus Autofluorescence for the Evaluation of Retinal Diseases
07:22

Quantitative Fundus Autofluorescence for the Evaluation of Retinal Diseases

Published on: March 11, 2016

11.8K
Retrospective Cardiac Gating with A Prototype Small-Animal X-ray Computed Tomograph
05:32

Retrospective Cardiac Gating with A Prototype Small-Animal X-ray Computed Tomograph

Published on: February 21, 2025

611

Area of Science:

  • Ophthalmology
  • Medical Imaging
  • Artificial Intelligence

Background:

  • Low-quality fundus photographs with artifacts can lead to misdiagnoses in clinical practice.
  • Cycle-consistent Generative Adversarial Network (CycleGAN) models can generate images without paired data.

Purpose of the Study:

  • To present a deep learning technique for automatic artifact removal in fundus photographs using a CycleGAN model.
  • To evaluate the effectiveness of CycleGAN in enhancing the quality of retinal images.

Main Methods:

  • A CycleGAN model was applied to 2206 anonymized color fundus photographs (256x256x3 resolution).
  • The dataset was divided into training (90%) and testing (10%) sets for model application.
  • Automated Quality Evaluation (AQE) was used to assess image quality before and after artifact removal.

Main Results:

  • CycleGAN successfully reduced artifacts like haze, lashes, and uneven illumination in generated images.
  • The core retinal information was preserved in the enhanced images.
  • Generated images showed improved AQE grade values compared to original images with artifacts.

Conclusions:

  • CycleGAN effectively reduces artifacts and improves fundus photograph quality, aiding clinical analysis.
  • The technique offers potential benefits for interpreting low-quality retinal images.
  • Future research should focus on enhancing the resolution and detail of generated images.