Detection of neovascularization in retinal images using mutual information maximization

Gradation of diabetic retinopathy on reconstructed image using compressed sensing

“…Therefore, the V d and the V t of the thin vessel class are computed and their MI is maximised using DE in 2D feature space to obtain the optimal threshold values (say T 1 and T 2 , respectively). Vessels with V d > T 1 and V t > T 2 are considered to be abnormal and are logically OR fusioned to obtain the entire neovascularisation structure, as reported in [14].…”

“…The main objective of the present work is to develop an integrated system design platform that efficiently detects the abnormal thin vessels (neovascularisation) and the lesions for gradation of DR under CS framework, to the best of our knowledge, which is not yet reported in the existing literature. A two-stage classification [14] where a two-level mutual information (MI) maximisation scheme is used for vessel extraction and neovascularisation detection. The image reconstruction using CS enhances the uncertainty on different types of blood vessels' classification.…”

A cybernetic systems approach to abnormality detection in retina images using case based reasoning

“…al. [4], DR in advanced stage called proliferative stage (PDR) or neovascularisation is considered and the development of new abnormal blood vessels, is detected. But in our current research work, the different signs of non-proliferative stages of DR like MA, HE and the starting signs of AMD are identified along with some proliferative stage abnormalities of DR like HAM.…”

The computer based method to diabetic retinopathy assessment in retinal images: a review

“…They used curvelet transform to enhance the tiny blood vessels then they applied maximization of mutual information (MI) on the maximum matched filter response for finding of optimal thresholding to classification of the vessels into the thick and thin class. Vessel density and tortuosity are computed from the thin vessel class followed by MI maximization and post-processing for neovascularization detection (66). Huang et al presented an automated system for detection NVE by using extreme learning machine.…”