Title :
The Optimisation of Thresholding Techniques for the Identification of Choroidal Neovascular Membranes in Exudative Age-Related Macular Degeneration
Author :
Brankin, E. ; McCullagh, Paul ; Black, N. ; Patton, W. ; Muldrew, A.
Author_Institution :
Fac. of Eng., Ulster Univ., Jordanstown
Abstract :
The application of image processing to the investigation of age-related macular degeneration (AMD) has focused on detecting focal drusen deposits in retinal images. This research investigates algorithmic approaches in order to detect choroidal neovascularisation (CNV) from retinal fluorescein angiograms in exudative AMD, the most severe form of the disease. A combination of the ´Sobel´ edge detection algorithm combined with thresholding produced the best qualitative segmentation, as verified by a trained ophthalmic grader. This study confirms that image processing can be used to identify certain types of CNV in retinal images particularly those that are hyper fluorescent. Further work is necessary to quantify the total lesion and characterise the clinically significant sub-components: classic or occult leakage, blood or exudate
Keywords :
biomedical optical imaging; biomembranes; diseases; dyes; edge detection; eye; image segmentation; medical image processing; optimisation; Sobel edge detection algorithm; blood; choroidal neovascular membranes; choroidal neovascularisation; classic leakage; exudate; exudative age-related macular degeneration; focal drusen deposits; image processing; occult leakage; optimisation; qualitative segmentation; retinal fluorescein angiograms; retinal images; thresholding techniques; Biomedical imaging; Biomembranes; Blindness; Blood vessels; Image edge detection; Image processing; Lesions; Medical treatment; Pigmentation; Retina;
Conference_Titel :
Computer-Based Medical Systems, 2006. CBMS 2006. 19th IEEE International Symposium on
Conference_Location :
Salt Lake City, UT
Print_ISBN :
0-7695-2517-1
DOI :
10.1109/CBMS.2006.157
