Related Experiment Video
Updated: May 25, 2026
09:17
Using Retinal Imaging to Study Dementia
Published on: November 6, 2017
Improving vessel segmentation in ultra-wide field-of-view retinal fluorescein angiograms
A Perez-Rovira1, K Zutis, J P Hubschman
1School of Computing, University of Dundee, Scotland. arovirez@computing.dundee.ac.uk
Summary
This study presents a new method for retinal vessel segmentation in ultra-wide field fluorescein angiograms. The technique improves accuracy by penalizing false positives, enhancing peripheral vessel detection.
Area of Science:
- Ophthalmology
- Medical Imaging
- Computer Vision
Background:
- Retinal vessel segmentation is crucial for diagnosing eye diseases.
- Ultra-wide field fluorescein angiography (UWFA) presents unique challenges due to dynamic perfusion and varying vessel appearance.
Purpose of the Study:
- To develop and evaluate an improved method for retinal vessel segmentation in UWFA sequences.
- To enhance the accuracy and robustness of vessel detection, particularly in challenging peripheral regions.
Main Methods:
- Utilized steerable filters and automatic thresholding for initial vessel segmentation.
- Introduced a novel penalization stage to refine segmentation by reducing false positives around the optic disc, in choroidal vessels, and lesions.
Main Results:
- The penalization stage significantly improved segmentation performance.
- Achieved an 11.84% increase in precision, a 12.98% increase in recall, and a 0.40% increase in accuracy.
- Demonstrated enhanced detection of peripheral vessels and reduced false positives.
Conclusions:
- The proposed method offers a robust solution for retinal vessel segmentation in UWFA.
- The availability of the algorithm, dataset, and results promotes further research and clinical application.
