Abstract: Diabetes Mellitus, also known as diabetes, causes persistently high blood sugar levels that can lead to diabetic retinopathy, which has been linked to damage to the retina's tiny blood vessels. The retina must first register light before the optic nerve can transmit signals to the brain. Until diabetic retinopathy begins to advance towards Proliferate DR/PDR, treatment is frequently postponed. As Diabetic Retinopathy (DR) worsens, more regular comprehensive dilated eye exams are required. Severe non-proliferative diabetic retinopathy patients are at a high risk of developing PDR and may require a thorough dilated eye exam every two to four months. [1] Therefore, in our work, we constructed a model called "retina.model" that can recognise even the slightest variation between each stage of DR and is 100% reusable with a growing amount of cognition over time as the computer tries to learn new patterns.

Keywords: matrix handling, Diabetes , American Optometric Association (AOA), Deep Learning, CNN architecture, Diabetic Retinopathy (DR), Image Classification, retina of the eye, Optometrist, Gaussian filters, Mild DR and Moderate DR, "retina.model", Severe non-proliferative diabetic retinopathy and Cotton wool spots.


PDF | DOI: 10.17148/IJARCCE.2023.125163

Open chat
Chat with IJARCCE