| Sumario: | In this modern world, diagnosis of disease by using advanced technology plays a vital role in the medical field. Medical imaging techniques are used to detect diseases by visualizing the image. A better way to detect the disease is using Deep learning (DL) algorithm. DL is the subset of artificial intelligence (AI) which thus a subset of computerized reasoning(AI). In deep learning technique, the features are picked out by the neural network without human intervention. Diabetic retinopathy is caused by high sugar level in the blood which damages the back part of the eye called retina. Ineffectively controlled glucose is one of the danger factors to cause diabetic retinopathy. Early symptoms include floaters, blurriness, micro-aneurysms and difficulty in perceiving colours. In severe cases, blindness can occur. Due to this reason, the early detection of retinopathy is of critical importance. The retinal fundus photographs will be processed with the help of a deep learning algorithm called "Convolutional Neural Network" which helps in classifying the image by learning distinctive features of the retinal image by itself. Diabetic retinopathy can be recognized a lot quicker and more exact outcomes will be normal with the assistance of CNN calculation. Since, it is more efficient by identifying the features accurately as well as its relation with other features in an image. This paper proposes the solution for diabetic retinopathy detection in eyes using deep learning.
|
|---|