WSEAS Transactions on Biology and Biomedicine
Print ISSN: 1109-9518, E-ISSN: 2224-2902
Volume 12, 2015
Automatic Classification of the Severity Level of the Retinal Ischemia
Authors: , , , , ,
Abstract: When the blood vessels stop supplying any region in the retina of the eye, this region is called capillary non-perfusion (CNP). With increasing and spreading of these regions across the retina, the patient can go blind. These regions appear only in the fundus fluorescein angiograms (FFAs). In this paper, an algorithm to automate the segmentation and classification of these regions is presented. The segmentation algorithm consists of three main steps: pre-processing, vessels extraction and CNP segmentation. After that, the automatic classification is applied to determine the severity level of each image. In the segmented algorithm, the CNPs are extracted by using the region growing algorithm. The algorithm is tested on 88 FFA images and evaluated by using two different ground truth images. The severity level is classified based on the percentage of the CNPs in each image.
Search Articles
Keywords: Retinal ischemia, Classification, Capillary non-perfusion, Ischemia segmentation, region growing
Pages: 44-50
WSEAS Transactions on Biology and Biomedicine, ISSN / E-ISSN: 1109-9518 / 2224-2902, Volume 12, 2015, Art. #7