Retinal vessel extraction by combining radial symmetry transform and iterated graph cuts

Area of Science:

• Ophthalmology
• Medical Imaging
• Computer Vision

Background:

• Accurate segmentation of retinal blood vessels is crucial for diagnosing various ocular diseases.
• Existing methods often struggle with noise and low contrast in retinal images, impacting diagnostic accuracy.

Purpose of the Study:

• To develop and evaluate a novel, robust method for automated retinal blood vessel extraction.
• To improve the accuracy and reliability of retinal image analysis for clinical applications.

Main Methods:

• A Hessian-based multiscale filtering technique was employed to enhance retinal blood vessel structures.
• A novel radial symmetry transformation using line kernels was introduced to improve vessel detection and suppress non-vessel structures.
• An iterated segmentation algorithm was utilized for the final extraction of retinal vessels.

Main Results:

• The proposed method demonstrated effective enhancement of blood vessels in grayscale retinal images.
• The radial symmetry transformation successfully improved vessel structure detection while reducing false positives from non-vessel elements.
• Validation on the DRIVE and STARE datasets confirmed the feasibility and performance of the proposed retinal vessel extraction technique.

Conclusions:

• The developed method offers a promising approach for automated retinal blood vessel segmentation.
• This technique has the potential to aid in the early detection and monitoring of eye diseases through improved image analysis.
• The findings support the clinical utility of advanced image processing techniques in ophthalmology.