Retinal Vessel Radius Estimation and a Vessel Center Line Segmentation Method Based on Ridge Descriptors

Authors:
Changhua Wu;Jennifer J. Kang Derwent;Peter Stanchev
Affiliations:
Department of Computer Science, Kettering University, Flint, USA 48504;Department of Biomedical Engineering, Illinois Institute of Technology, Chicago, USA 60616-3793;Department of Computer Science, Kettering University, Flint, USA 48504
Venue:
Journal of Signal Processing Systems
Year:
2009

Citing 8
Cited 1

Edge Detection and Ridge Detection with Automatic Scale Selection

International Journal of Computer Vision
A Computational Method for Segmenting Topological Point-Sets and Application to Image Analysis

IEEE Transactions on Pattern Analysis and Machine Intelligence
Adaptive Local Thresholding by Verification-Based Multithreshold Probing with Application to Vessel Detection in Retinal Images

IEEE Transactions on Pattern Analysis and Machine Intelligence
Blood Vessels Segmentation in Retina: Preliminary Assessment of the Mathematical Morphology and of the Wavelet Transform Techniques

SIBGRAPI '01 Proceedings of the 14th Brazilian Symposium on Computer Graphics and Image Processing
Muliscale Vessel Enhancement Filtering

MICCAI '98 Proceedings of the First International Conference on Medical Image Computing and Computer-Assisted Intervention
Rapid automated tracing and feature extraction from retinal fundus images using direct exploratory algorithms

IEEE Transactions on Information Technology in Biomedicine
Robust model-based vasculature detection in noisy biomedical images

IEEE Transactions on Information Technology in Biomedicine
Segmentation of vessel-like patterns using mathematical morphology and curvature evaluation

IEEE Transactions on Image Processing

Retinal vessel segmentation using a multi-scale medialness function

Computers in Biology and Medicine

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper studies the retinal vessel radius estimation and proposes a segmentation method for vessel center lines based on ridge descriptors. The study on radius estimation reveals that the radius estimation by the matched filters based on the second order derivatives of Gaussian kernels is only correct at the vessel center. The relation between the vessel radius and the scale of the Gaussian kernel in the estimation method based on the normalized largest curvature is also studied. The ridge descriptor proposed in this paper contains the normalized largest curvature and the orientations of gradients in the local neighborhood. For vessels of a certain scale, the distribution of the descriptors is assumed to be a normal distribution and is learned from a training set with known truth. Vessel center line segmentation can be then performed based on the distance between the ridge descriptor at candidate pixels and the learned model. Evaluation of the vessel center line segmentation based on the descriptors is done on both DRIVE and STARE databases using the receiver operating characteristic (ROC) curves. The areas under the ROC curves on DRIVE and STARE databases are 0.9584 and 0.9421 respectively.