Detection of Retinal Neovascularization Using Optimized Deep Convolutional Neural Networks

The most common disease that is found among people across the world Diabetes and it predicted to increase more in upcoming years by World Health Organization (WHO). People who are diabetic for a longer period likely have Diabetic Retinopathy (DR), an eye which can lead blindness this cannot be reversed. One of severe stage problems DR Retinal Neovascularization (RN), i.e., outburst retinal blood vessels. Residual Network (ResNet) has effective technique called Skip or Connections solves problem vanishing gradient during backpropagation. ResNet50 50 layers deep network omits signal representations learns from residual leading predict RN with 88.97% accuracy.