Developing an adaptive neuro fuzzy inference system anfis classifier to diagnose diabetic retinopathy in fundus images

Abstract

Diabetes disease becomes common among human beings globally newlineand it affects people irrespective of age. Due to the death rate, disability and newlineblindness, the diabetic disease has been marked as a leading one among all newlineother diseases in the world. Commonly, the diabetic people are affected by newlineretinopathy problem which is the most common eye disease. Diabetic newlineRetinopathy is the most widely recognized reason for visual deficiency for the newlinediabetic patients before 50 years of age. The main problem associated with newlinethe Diabetic Retinopathy is that the patient is not aware of the warning newlinesymptoms or indications at the early phase of disease. The level and severity newlineof this disease can be pathological identified by types and number of lesions newlinepresent in the surface of the retina. Diabetic Retinopathy is extensively newlinesegregated into two major types namely Non Proliferative Diabetic newlineRetinopathy (NPDR) and Proliferative Diabetic Retinopathy (PDR). Among newlinethe various problems of NPDR, exudates in the retina can be identified and it newlinehas a possible treatment at its early stage. But the treatment is impossible newlinewhen the level is at essential stage. The presence of exudates lesion indicates newlineharm to the retinal micro blood vessels and in spite of the fact that the newlinepathogenic instruments are not completely comprehended, the clinical bio newlinefeatures remains distinctive and effortlessly perceived. newlineTraditional Diabetic Retinopathy diagnosis models are found to be newlineexpensive, time consuming, and requires experienced professionals. To newlineresolve these problems, an alternative method of Diabetic Retinopathy newlinediagnosis can be designed using machine learning (ML) and computer vision newlineapproaches newline