|
KMID : 1147120100160010031
|
|
Journal of the Korean Society of Imaging Informatics in Medicine
2010 Volume.16 No. 1 p.31 ~ p.36
|
|
|
Automated Macular Drusen Detection from Fundus Images Using Mean Filters
|
|
Kim Young-Jae
Kim Kwang-Gi
Nam Kyoung-Won
Jeong Chang-Bu
Kim Yu-Shin
Yang Hee-Kyung
Ahn Jee-Yun
Hwang Jeong-Min
|
|
|
Abstract
|
|
|
|
Age-related macular degeneration (ARMD) is a degenerative disease characterized by a progressive damage to a yellow spot, the center of retina. Drusen are tiny yellow or white extracellular material accumulated between the retinal pigment epithelium (RPE) and the underlying choroid. They cause macular degeneration and could result in a loss of vision in severe cases. Thus, the accurate measurement of drusen, especially hard ones, is important for proper treatment of degenerative macular diseases. However, manual detection of drusen by clinicians is a time consuming and difficult process. In this study, an algorithm for the automatic detection of drusen was developed using imageprocessing techniques, a mean filter and thresholding. The algorithm was tested using images of five ARMD patients. Results showed that the sensitivity of the automatic drusen detection was 84.2~97.8% (mean = 92.7%). The proposed algorithm for automatic drusen detection is expected to effectively assist ophthalmologists in the diagnosis of drusen.
|
|
|
KEYWORD
|
|
|
ARMD, drusen, mean filter, threshold
|
|
|
FullTexts / Linksout information
|
|
|
|
|
|
Listed journal information
|
|
|
|