Related Experiment Video
Updated: Apr 27, 2026
12:28
Quantification of Vascular Parameters in Whole Mount Retinas of Mice with Non-Proliferative and Proliferative Retinopathies
Published on: March 12, 2022
3.3K
Level Sets for Retinal Vasculature Segmentation Using Seeds from Ridges and Edges from Phase Maps
Bekir Dizdaroğlu1, Esra Ataer-Cansizoglu2, Jayashree Kalpathy-Cramer3
1Computer Engineering Department, Karadeniz Technical University, Turkey ; Cognitive Systems Laboratory, Northeastern University, Boston MA, USA.
IEEE International Workshop on Machine Learning for Signal Processing : [Proceedings]. IEEE International Workshop on Machine Learning for Signal Processing
|July 1, 2014
Summary
This study introduces a new automated method for retinal vasculature segmentation using ridge sampling and phase-based edge detection. The novel approach accurately segments blood vessels in fundus images, improving upon existing level set techniques.
Area of Science:
- Medical image analysis
- Computer vision
- Ophthalmology
Background:
- Accurate retinal vasculature segmentation is crucial for diagnosing eye diseases.
- Traditional level set methods struggle with accurate boundary detection and automation in fundus images.
Purpose of the Study:
- To develop a novel, automated level set-based method for retinal vasculature segmentation.
- To improve accuracy and overcome limitations of existing segmentation techniques.
Main Methods:
- Introduced ridge sample extraction for centerline sampling.
- Implemented phase map-based edge detection for precise boundary identification.
- Utilized ridge identification for automated seed point initialization.
Main Results:
- The proposed method achieves accurate and automatic segmentation of retinal vasculature.
- Quantitative and visual results demonstrate superior performance compared to classical methods.
- The algorithm successfully segments vessels in challenging fundus imagery.
Conclusions:
- The novel modification enhances level set-based retinal vasculature segmentation.
- The approach offers a robust and automated solution for medical image analytics.
- This methodology is applicable to other vessel segmentation challenges in medical imaging.
