A Review of Image Processing and Deep Learning Based Methods for Automated Analysis of Digital Retinal Fundus Images

Retinal fundus imaging is a medical procedure used by medical professionals in the discovery and tracking of various retinal abnormalities. Sometimes the analysis of retinal fundus images can be slow and difficult when performed by medical staff, and in response to this many automated, image-processing based methods for the analysis of these images exist. In recent years, deep learning methods have become increasingly popular in machine learning applications, so it is no surprise that they are also being used in the image processing based analysis of retinal fundus images. In this paper we discuss recently proposed methods that use deep learning techniques in the image processing based analysis of digital retinal fundus images. Special attention is given to the analysis of retinal fundus image datasets and various techniques employed to the images from these datasets in order to make them suitable for deep learning based applications.