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 Videos

Automated microaneurysm detection using local contrast normalization and local vessel detection.

Alan D Fleming1, Sam Philip, Keith A Goatman

  • 1Biomedical Physics, University of Aberdeen, Aberdeen, AB25 2ZD, UK. a.fleming@biomed.abdn.ac.uk

IEEE Transactions on Medical Imaging
|September 14, 2006
PubMed
Summary
This summary is machine-generated.

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

Podcasts as Digital Scholarship in Surgical Education and Professional Development: Perspectives from Learners to Leaders.

Annals of surgery·2026
Same author

Profiling the CFTR Variant Selectivity and Off-Target Interactions of VX-121.

bioRxiv : the preprint server for biology·2026
Same author

Spatial Mapping of the Precancer-to-Cancer Transition in Breast and Prostate.

Cancer discovery·2026
Same author

General trends in the calnexin-dependent expression and pharmacological rescue of clinical CFTR variants.

eLife·2025
Same author

Awareness and Perceptions of Flexibility in Surgical Training: Insights From a National Survey of Surgical Fellowship Program Directors.

Journal of surgical education·2025
Same author

Impact of Surgeon-Scientist Institutional Transitions on Funding Progression.

Annals of surgery·2025

Automated analysis of retinal images can detect diabetic eye disease early. Image contrast normalization improves microaneurysm detection, aiding in classifying retinopathy signs.

Area of Science:

  • Ophthalmology
  • Medical Imaging
  • Computer Vision

Background:

  • Diabetic eye disease screening programs utilize retinal photography.
  • Automated image grading is being explored to reduce human workload.
  • Microaneurysms (MAs) are critical early indicators of diabetic retinopathy.

Purpose of the Study:

  • To develop and evaluate automatic methods for microaneurysm (MA) detection in retinal images.
  • To assess the impact of image contrast normalization on MA detection accuracy.
  • To compare different contrast normalization techniques for improved lesion identification.

Main Methods:

  • Implementation of automatic methods for microaneurysm detection.
  • Application and comparison of various image contrast normalization techniques.

Related Experiment Videos

  • Utilizing watershed transform for vessel and lesion exclusion, and local vessel detection for handling dots within vessels.

    • Main Results:

    • Contrast normalization significantly improves the distinction between MAs and other retinal dots.
    • The best contrast normalization method involved watershed transform for region derivation.
    • Detection of images containing MAs achieved 85.4% sensitivity and 83.1% specificity.

    Conclusions:

    • Automated MA detection methods, enhanced by contrast normalization, are effective for diabetic retinopathy screening.
    • The proposed techniques can aid in reducing the burden on human graders for retinal image analysis.
    • Accurate detection of MAs is crucial for classifying diabetic eye disease from retinal photographs.