Automatic Diabetic Retinopathy Detection Through Ensemble Classification Techniques Automated Diabetic Retionapthy Classification

Retinal image analysis is one of the profound areas of research. Computational solutions are sought in this context. Data Mining techniques have been extensively adopted for this purpose. In this work, Diabetic Retinopathy, a primary cause of blindness is dealt with. A two-level classification is adopted to classify Diabetic Retinopathy. In this classification, first level classification is performed through ensemble of Best First Trees. The misclassified instances are considered as noise and outlier and removed. The resultant clean data is then subjected to second level classification, performed through ensemble of J48Graft trees. The evolved rules are evaluated through 3 fold cross validation and the best rules are generated for detection of Diabetic Retinopathy. The proposed methodology is tested on Diabetic Retinopathy Drebechen Dataset available from UCI Machine Learning Repository. The method achieves an accuracy of 96.14% revealing a marked improvement when compared to the earlier works. The system can be of great help to the Ophthalmologists society.