Diabetic Retinopathy Detection Using Ensemble of CNN Architectures

Diabetic retinopathy (DR) is an eye disease that affects diabetes patients and damages their retina if not detected early. DR has been manually screened by an ophthalmologist until recently, which is a time-consuming technique. The objective is to use AI technology to automatically scan photographs for the disease and provide information on the severity of the issue. We propose a deep learning model that can automatically analyze a patient’s ocular picture and determine the severity of blindness faster. Based on this blindness severity, we can screen the process of treating DR on a wide scale with pace. The proposed computationally efficient ensemble of CNN architectures is used to correctly detect and classify diabetic retinopathy. This model is less complicated, and it results in 72 \(\%\) of accuracy tested on retina fundus image dataset.