Multi-level Quality Assessment of Retinal Fundus Images using Deep Convolution Neural Networks

Retinal fundus image quality assessment is one of the major steps in screening for retinal diseases, sincethe poor-quality retinal images do not allow an accurate medical diagnosis. In this paper, we first introducea large multi-level Retinal Fundus Image Quality Assessment (RFIQA) dataset. It has six levels of qualitygrades, which are based on important regions to consider for diagnosing diabetic retinopathy (DR), AgedMacular Degeneration (AMD) and Glaucoma by ophthalmologists. Second, we propose a Convolution NeuralNetwork (CNN) model to assess the quality of the retinal images with much fewer parameters than existingdeep CNN models and finally we propose to combine deep and generic texture features, and using RandomForest classifier. Experiments show that combing both deep and generic features outperforms using any of thetwo feature types in isolation. This is confirmed on our new dataset as well as on other public datasets