Diabetic Retinopathy Identification From Retinal Images With The Use Of A Deep Learning-Based Transfer Training Framework

Diabetic retinopathy (DR) is a major contributor to vision loss worldwide, highlighting the urgent need for early identification and prompt care to prevent further deterioration of vision. Diagnosing DR is intrinsically difficult, since it requires the thorough evaluation of complex retinal images by skilled professionals. DR, a consequence of diabetes, has the potential to cause vision loss if not identified in its early stages. research presents a novel model that utilizes transfer learning to detect diabetic retinopathy. The first trials yielded poor accuracies for both training and validation. The categorization of class-4, especially those depicting severe instances, yielded unexpected outcomes, since several images were categorized as moderate. Consequently, more investigation is required to ascertain the accuracy of these classifications. Therefore, our suggested approach allows for easy classification of the retina pictures of both Diabetic and Healthy individuals, resulting in a decrease in the number of evaluations required by medical specialists.