Optimized Diabetic Retinopathy Eye Disease Classification Using Deep Learning Model

Globally, retinal diseases-including diabetic retinopathy and other degenerative conditions are the major leading causes of blindness and visual impairment. Even with major improvements in treatment choices, long-term vision deterioration and higher healthcare expenses are frequently the result of inadequate early detection systems. Irreparable harm can be avoided with an early and precise diagnosis of retinopathy; yet, manual retinal image analysis is still laborious and subject to human error. This demonstrates the pressing need for sophisticated automated systems capable of accurately dividing retinal diseases into healthy and unhealthy categories. By utilizing state-of-the-art image classification methods, the suggested approach seeks to improve diagnostic accuracy and help physicians spot retinal disorders early. By tackling present diagnostic difficulties, this strategy could lower the number of blind people, ease the strain on medical professionals, and enhance patient outcomes in general. The MobileNetV2 algorithm demonstrated the performance, by achieving an accuracy of 95%, compared to Mobile Netaccuracy of 80%.