A Study of Performance Evaluation of Convolution Neural Network for Diabetic Retinopathy

Kusuma R C
2020 International Journal of Advanced Trends in Computer Science and Engineering  
Diabetic retinopathy is a retinal illness, caused by high blood sugar level, which will lead to damage in retinal blood vessels. It is difficult to detect the disease in early stages, as it will not show any symptoms at the early stage. DR is the major reason for the blindness in Diabetic patients. Clinicians recommends that the patient's needs to be diagnosed at least two times per year in order to identify signs of disease caused by diabetic retinopathy. At present diagnosis for the detection
more » ... of the diabetic retinopathy mainly depends on ophthalmologist who examines the retinal image and decides the patient's condition that they have diabetic retinopathy or not. Computer-aided verification of fundus images is necessary because it permits for fast processing and bunch images can be analyzed in one shot at a time. Consequently, to attain fast, reliable diagnosis and to treat the patient at the early stages application of deep CNNs approaches are used. Automatic diabetic retinopathy detection examines
doi:10.30534/ijatcse/2020/118942020 fatcat:errcchlo4bdijhzybuy4h44yhm