MRF reconstruction of retinal images for the optic disc segmentation

Authors:
Ana Salazar-Gonzalez;Yongmin Li;Djibril Kaba
Affiliations:
Department of Information Systems and Computer Science, Brunel Universit, London, UK;Department of Information Systems and Computer Science, Brunel Universit, London, UK;Department of Information Systems and Computer Science, Brunel Universit, London, UK
Venue:
HIS'12 Proceedings of the First international conference on Health Information Science
Year:
2012

Citing 5
Cited 0

Texture Synthesis by Non-Parametric Sampling

ICCV '99 Proceedings of the International Conference on Computer Vision-Volume 2 - Volume 2
Graph Cuts and Efficient N-D Image Segmentation

International Journal of Computer Vision
Topology cuts: A novel min-cut/max-flow algorithm for topology preserving segmentation in N-D images

Computer Vision and Image Understanding
Graph Cuts Segmentation with Statistical Shape Priors for Medical Images

SITIS '07 Proceedings of the 2007 Third International IEEE Conference on Signal-Image Technologies and Internet-Based System
Segmentation of the optic disk in color eye fundus images using an adaptive morphological approach

Computers in Biology and Medicine

Quantified Score

Hi-index	0.00

Visualization

Abstract

The retinal image analysis has been of great interest because of its efficiency and reliability for optical diagnosis. Different techniques have been designed for the segmentation of the eye structures and lesions. In this paper we present an unsupervised method for the segmentation of the optic disc. Blood vessels represent the main obstruction in the optic disc segmentation process. We made use of our previous work in blood vessel segmentation to perform an image reconstruction using the Markov Random Field formulation (MRF). As a result the optic disc appears as a well defined structure. A traditional graph is then constructed using spatial pixel connections as boundary term and the likelihood of the pixels belonging to the foreground and background seeds as regional term. Our algorithm was implemented and tested on two public data sets, DIARETDB1 and DRIVE. The results are evaluated and compared with other methods in the literature.