A NEW HYBRID ALGORITHM FOR RETINAL VESSELS SEGMENTATION ON FUNDUS IMAGES

Abstract

Automatic retinal vessels segmentation is an important task in medical applications. However, most of the available retinal vessels segmentation methods are prone to poorer results when dealing with challenging situations such as detecting low-contrast micro-vessels, vessels with central reflex and vessels in the presence of pathologies. This paper presents a new hybrid algorithm for retinal vessels segmentation on fundus images. The proposed algorithm overcomes the difficulty when dealing with the challenging situations by first applying a new directionally sensitive blood vessel enhancement method before sending fundus images to a convolutional neural network architecture derived from U-Net. To train and test the algorithm, fundus images from the DRIVE and STARE databases, as well as high-resolution fundus images from the HRF database, are utilized. In the experiment, the proposed algorithm outperforms state-of-the-art methods in four major measures, i.e. sensitivity, F1-score, G-mean and Mathews correlation coefficient both on the low and high-resolution images. In addition, the proposed algorithm achieves the best connectivity area- length score among the competing methods. Given such performance, the proposed algorithm can be adapted for vessel-like structures segmentation in other medical applications. In addition, since the new blood vessel enhancement method is independent of the U-Net model, it can be easily applied to other deep learning architectures. 

Existing System

Structured output support vector machine (SOSVM), conditional random field (CRF).

Proposed System

The proposed blood vessel segmentation algorithm combines a new blood vessel enhancement method and a CNN structure, derived from U-Net. The enhancement method is based on the multi-scale and orientation modified Dolph-Chebyshev type I function (MDCF-I) matched filter. The matched filter is modified such that it can be used to detect blood vessels with all possible calibres. The segmentation output obtained using the proposed algorithm is also evaluated under the connectivity area-length (f (C; A;L)) score. This score measures the amount of vessel tree features preserved in the segmentation output. In addition, since the proposed enhancement method is independent of deep learning architectures, it can be easily applied to other deep learning models.

Architecture

BLOCK DIAGRAM