Open AccessMachine Learning Technique for Feature Extraction and Segmentation of Retinal Blood Vessels
Author(s) -
Shivani A Patil MTECH,
Prof. Dr. Pradnya Kulkarni
Publication year - 2020
Publication title -
international journal of recent technology and engineering
Language(s) - English
Resource type - Journals
ISSN - 2277-3878
DOI - 10.35940/ijrte.a2223.059120
Subject(s) - fundus (uterus) , diabetic retinopathy , computer science , retinal , artificial intelligence , segmentation , retina , blindness , ophthalmology , computer vision , optometry , feature extraction , medicine , retinopathy , pixel , feature (linguistics) , diabetes mellitus , neuroscience , psychology , linguistics , philosophy , endocrinology
The term Diabetic retinopathy is the serious issue that is caused by the diabetes, which affects the eyes that may lead to blindness. DR takes place due to damaged arteries and veins that are a part of the fundus of the eye. Although DR can be prevalent now days, its prevention remains challenging. Visual analysis of the funds and consideration of colour photographs Ophthalmologists directly examine the existence and severity of DR. This process is expensive as well as time consuming as there are huge number of diabetes affected people worldwide. The automatic Diabetic Retinopathy system is expanded to predict various related diseases that are analysed. The proposed methodology uses EYEPACS dataset that consists on 35,126 fundus images. These images are pre-processed and sent to neural network that detects type of DR. CNN detects clusters of pixels that are damaged in the macula region and in turn evaluates the overall damaged area in the macula from the retinal images. The retinal fundus images present structural and impulsive noise.