Automated detection of AMD using wavelet transform based features

Macula is the region in the human retina that takes care of central vision. Many people aged above 50 years develop degeneration of the macular region as a part of the body's natural aging process. Age related macular degeneration is the most common type of macular problem that may lead to blindness. In this paper, we present a method of detecting drusen in retinal fundus images. Drusen is detected by using two methods. In the first method drusen is detected by eliminating background and in the second method it is detected by eliminating blood vessels. Result of the two methods is added for final detection of drusen. Two dimensional wavelet transform is applied and then wavelet based features are extracted. These features are used for classification using SVM classifier. Highest values achieved are: 93.3% Sensitivity, 100% Specificity and 95% Accuracy.