A novel vessel segmentation algorithm for pathological retina images based on the divergence of vector fields

Area of Science:

• Ophthalmology
• Medical Imaging
• Computer Vision

Background:

• Accurate detection of retinal blood vessels is crucial for diagnosing various eye conditions.
• Pathological changes in retinal images present challenges for automated vessel segmentation.
• Existing methods may struggle with noise and false vessel detection in diseased retinal regions.

Purpose of the Study:

• To propose and evaluate a novel method for robust blood vessel detection in pathological retina images.
• To improve the accuracy of retinal vessel segmentation, particularly in challenging pathological areas.
• To provide a reliable tool for analyzing retinal vasculature in both healthy and diseased states.

Main Methods:

• Blood vessel-like objects are extracted using the Laplacian operator.
• Centerlines are detected using the normalized gradient vector field for noise pruning.
• The proposed method was validated on the publicly available STARE database of pathological retina images.

Main Results:

• The method successfully avoids the detection of false vessels in pathological regions of the retina.
• Reliable blood vessel detection results were achieved even in the presence of image pathology.
• The approach demonstrated effectiveness across a comprehensive set of pathological retina images.

Conclusions:

• The proposed method offers a significant improvement for blood vessel detection in pathological retinal images.
• This technique provides a reliable and accurate means for analyzing retinal vasculature, aiding in clinical diagnosis.
• The method's ability to handle pathological conditions makes it a valuable tool for ophthalmological research.