Segmentation of retina images to extract abnormalities, and in particular, exudates, using fuzzy c-means have been
used successfully in earlier work. Mathematical morphology has also been used to extract exudates and micro
aneurysms 9, 13. Hue and intensity characteristics of regions of interest have been utilized in exudates extraction 10.
Naive Bayes and SVM classification has been used to classify exudates 14. Micro aneurysms have been detected
using Gabor filters 15. Decision Support systems have been used to classify diabetic retinopathy images 12. Machine
learning has been used for exudates extraction 14. Decision trees along with association rule mining has been used
for general medical image classification 17. A comparative study of Case Based Reasoning and data mining has been
carried out 16. These approaches have been confined to DR cases. In our proposed algorithm, the Decision Tree
approach has been employed to integrate AMD cases with DR. The advantage is in the use of contextual
information to differentiate between AMD and DR.