AUTOMATED METHOD FOR RETINAL ARTERY/VEIN SEPARATION VIA GRAPH SEARCH METAHEURISTIC APPROACH

Abstract

Separation of the vascular tree into arteries and veins is a fundamental prerequisite in the automatic diagnosis of retinal biomarkers associated with systemic and neurodegenerative diseases. In this paper, we present a novel graph search metaheuristic approach for automatic separation of arteries/veins (A/V) from color fundus images. Our method exploits local information to disentangle the complex vascular tree into multiple subtrees, and global information to label these vessel subtrees into arteries and veins. Given a binary vessel map, a graph representation of the vascular network is constructed representing the topological and spatial connectivity of the vascular structures. Based on the anatomical uniqueness at vessel crossing and branching points, the vascular tree is split into multiple subtrees containing arteries and veins. Finally, the identified vessel subtrees are labeled with A/V based on a set of handcrafted features trained with random forest classifier. The proposed method has been tested on four different publicly available retinal datasets with an average accuracy of 94.7%, 93.2%, 96.8% and 90.2% across AV-DRIVE, CT-DRIVE. INSPIRE-AVR and WIDE datasets, respectively. These results demonstrate the superiority of our proposed approach in outperforming state-of-the-art methods for A/V separation.

Existing System

Graph search optimization, genetic search based feature selection technique

Proposed System

In this paper, we propose a novel graph search metaheuristic approach for automatic A/V separation from retinal color fundus images. Here, we extend our previously proposed vessel keypoint detector (VKD) to incorporate the curvature characteristics of crossing vessel segments, along with the orientation and width information. This curvature property acts as a unique feature to VKD to aid in resolving the possible conflict, in assigning an A/V label of the highly curved crossover point. Besides, we present a novel graph search metaheuristic algorithm to generate anatomically meaningful vessel subtrees by searching the space of possible connectivity of vascular networks. Our main contributions can be summarized as follows: 1) We propose an extended vessel keypoint detector which integrates curvature with the orientation and width information to precisely disentangle all crossing vessel pairs into corresponding A/V segments. 2) We propose a novel depth-first search based graph search metaheuristic algorithm to accurately identify all A/V vessel subtrees from a given vascular topology. 3) We extensively validated our method on four challenging publicly available retinal datasets, including images from two different imaging modalities - fundus as well as UWFov-SLO images.

Architecture

BLOCK DIAGRAM