Blood vessel segmentation in retinal fundus images using Gabor filters, fractional derivatives, and Expectation Maximization

Hugo Aguirre-Ramos, Juan Gabriel Avina-Cervantes, Ivan Cruz-Aceves, José Ruiz-Pinales and Sergio Ledesma

Applied Mathematics and Computation, 2018, vol. 339, issue C, 568-587

Abstract: In recent decades, the eye diseases have become the leading causes of blindness in young adults. Most of the cases can be prevented if detected in the early stages. For instance, the analysis of retinal blood vessels can help the physician to detect and prescribe appropriate treatment to the diabetic patient as a special case. This work describes a novel framework for blood vessels detection in retinal images. In the proposed methodology, the noise present in the green channel of the RGB image is reduced by a Low-Pass Radius Filter, subsequently, a 30-element Gabor filter and a Gaussian fractional derivative are used to remarkably enhance both the blood vessels structure and its contours. Thereafter, a threshold and a series of morphology-based decision rules are applied to isolate the blood vessels and reduce the incidence of false positive pixels. Additionally, our method can be used to detect the Optic Disc in the original image and remove it from the threshold result. The proposed method was assessed using the public DRIVE database, for the Test image set and the 1st manual delineations. In this database, our method is able to obtain an average accuracy of 0.9503, an average specificity of 0.7854, and an average balanced accuracy of 0.8758. Moreover, the proposed method shows a better performance than comparative methods, such as the threshold for a Frangi filter, Adaptive Threshold, and multiple classes Otsu method. After the analysis of the computer simulations, it was concluded that the proposed method is a competitive and reliable methodology for blood vessels segmentation.

Keywords: Segmentation; Gabor Filter; Fractional derivatives; Radius filter; Retinal fundus images (search for similar items in EconPapers)
Date: 2018
References: View complete reference list from CitEc
Citations: View citations in EconPapers (1)

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0096300318306234
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:apmaco:v:339:y:2018:i:c:p:568-587

DOI: 10.1016/j.amc.2018.07.057

Access Statistics for this article

Applied Mathematics and Computation is currently edited by Theodore Simos

More articles in Applied Mathematics and Computation from Elsevier
Bibliographic data for series maintained by Catherine Liu ().