Enhanced Detection of retinopathy affected blood vessels using deep convolutional neural networks
In human body, one of the complete sensory organ is eye. For getting the human vision perfectly, blood vessels in retina and neurons in eye plays a key role. Some of the diseases like hypertension, arteriosclerosis and diabetes retinopathy are causing the branch pattern change and also retinal blood vessels diameter leads to blindness. By segmenting the retinal blood vessels, we can analyze these changes. In this research work, we have proposed two types of neural network architecture based on deep learning technique which will segment the retina blood vessels. By pixel classification, the features map of different fundus images are retrieved by the multiple hidden layers. By the support of loss function, we are avoiding the losses due to the variation between the vessel and non-vessel. Proposed work will be examined for both of the architectures with database DRIVE.