Related Experiment Video
Updated: Jul 10, 2026
Using Retinal Imaging to Study Dementia
Published on: November 6, 2017
Detecting false vessel recognitions in retinal fundus analysis
Insights
This study introduces six methods to reduce false positives in retinal blood vessel tracking. These techniques analyze vessel properties along their full trajectory, improving diagnostic accuracy for eye diseases.
Area of Science:
- Medical Imaging
- Ophthalmology
- Computer Vision
Background:
- Automatic tracking of retinal blood vessels is crucial for diagnosing various eye diseases non-invasively.
- Current tracking methods often yield a high number of false positives, hindering accurate diagnosis.
Purpose of the Study:
- To develop and evaluate methods for discriminating false vessel detections from true positives in retinal fundus images.
- To improve the reliability of automated retinal vessel tracking.
Main Methods:
- Proposed six distinct methods, each modeling candidate vessels based on average geometric and grayscale properties along their entire trajectory.
- Utilized Fisher linear discriminant analysis for feature-based discrimination between true and false vessel detections.
- Leveraged information from the full vessel path to resolve ambiguities encountered during the tracking phase.
Main Results:
- Demonstrated satisfactory rejection of false positives across 28 retinal images.
- Observed improved performance with more complex vessel models.
- Successfully addressed limitations of small-scale tracking algorithms by analyzing full vessel trajectories.
Conclusions:
- The proposed methods effectively reduce false positives in retinal vessel tracking.
- Analyzing full vessel trajectory properties enhances the accuracy of automated diagnostic procedures.
- The approach offers a promising solution for more reliable non-invasive diagnosis of eye conditions.
Abstract:
Automatic tracking of blood vessels in images of retinal fundus is an important and non-invasive procedure for the diagnosis of many diseases. Tracking techniques often present a high rate of false positives. This paper presents six methods to discriminate false detections from true positives, each based on a different model of the vessel. They describe a candidate vessel in terms of its average geometric and grayscale properties considered along the full trajectory of the vessel itself. The rationale is that false vessels are caused by the small scale of the tracking algorithm necessary during the tracking phase. Once tracking has been completed, we can gather information from the full vessel trajectory and solve ambiguities that cannot be fixed during tracking. We apply Fisher linear discriminant analysis to these features to get the desired discrimination. Results on 28 images show satisfactory rejection of false positives and better results when using more complex models.
