A reliable automatic cataract detection using deep learning

An eye cataract is a serious condition in which the eye’s lens becomes clouding or less transparent and becomes the global cause of visual impairment. If the cataract is diagnosed accurately and on time, it can provide a significant benefit by increasing the life of the cataract patients. Hence, a convenient and cost-effective automatic method for cataract classification and diagnosis system is required. The main objective of the proposed study is to develop a fundus image analysis-based automatic classification and grading system. This study takes advantage of a deep convolution neural network (DCNN) to extract the features automatically from fundus images. Thereafter, these feature vector is applied to the soft-max function as a classifier to grade cataract severity into 4-stage namely mild, moderate, no, and severe. The reason for using fundus images is to capture the internal structure of the eyes accurately, which is needed in early medical diagnosis. The proposed approach achieved 92.7% accuracy for 4-stage cataract classification and grading, which is higher than the predictive values of other state-of-the-art algorithms.