Retinal blood vessel segmentation in diabetic retinopathy image using maximum tree

Faisal, Muhammad ORCID: https://orcid.org/0000-0003-4884-7254, Purnama, I Ketut Eddy, Hariadi, Mochamad ORCID: https://orcid.org/0000-0002-2194-169X and Purnomo, Mauridhie Hery (2012) Retinal blood vessel segmentation in diabetic retinopathy image using maximum tree. International Journal of Academic Research, 4 (3). pp. 83-88. ISSN 20757107

Preview

Text 15683.pdf - Published Version Download (3MB) | Preview

Abstract

Retinal blood vessels can give information about abnormalities or disease by examinate its pathological changes. Diabetic Retinopathy, disorder of retinal blood vessels resulting from diabetes mellitus, is one of major cause of human vision abnormalities or even blindness. Thus, segmentation of this feature in the retinal images can provide a map of retinal vessel can ease the assessment of the characteristics of the vessels. In this paper, a methods of blood vessel segmentation in Diabetic retinopathy image using attribute filtering which use Max-Tree to represents the image and branches filtering approach as its filtering process. Max-tree used to generate tree as image's representation based on its gray level. To determine which nodes preserved and removed, branches filtering is used which use leaf nodes as an initial reference in the filtering process This research uses 40 retinal images and its manual segmentation validated by expert observer included. Accuracy of this vessel segmentation method is 91,04% based on manual segmentation done by first expert observer and 92,19% based on second observer.

Item Type:	Journal Article
Keywords:	max-tree; branches filtering; fundus; retina, blood vessel
Subjects:	08 INFORMATION AND COMPUTING SCIENCES > 0801 Artificial Intelligence and Image Processing > 080106 Image Processing
Divisions:	Faculty of Technology > Department of Informatics Engineering
Depositing User:	Muhammad Faisal
Date Deposited:	02 Oct 2023 09:24

Downloads

Origin of downloads

Actions (login required)

View Item