Internet of Things Enabled Intelligent Machine Learning based Diabetic Retinopathy Grading and Classification Model

Abstract

Diabetic retinopathy (DR) is a commonly existing illness among diabetic patients and has become a major reason for blindness. The development of DR can be avoided by the earlier detection of the disease. The detection and severity estimation of DR can be made using retinal inspection. Besides, the machine learning (ML) models and Internet of Things (IoT) technologies help to detect and classify DR with enhanced performance and decreased cost. In this view, this paper presents an intelligent ML based DR grading and classification (IML-DRGC) model using retinal fundus images. The goal of the IML-DRGC model is to automatically diagnose the DR at maximum accuracy. The IML-DRGC model initially enables the IoT devices to capture the retinal fundus image of the patient. Then, the gathered image undergoes preprocessing using Gaussian filtering (GF) technique to get rid of the noise that exists in the fundus image. Followed by, fuzzy c-means enabled segmentation technique is exploited to detect the affected areas in the fundus image. Moreover, the Features from Accelerated Segment Test (FAST) model is utilized as a feature extractor. Finally, support vector machine (SVM) is used as a classification model to allot distinct labels of DR. For investigating the improved diagnostic outcome of the IML-DRGC model, a series of simulations were performed and the results are examined interms of different measures.