Fuzzy Logic Classifier Approach for Non-Proliferative Diabetic Retinopathy in Macular Disease Quantification and Diagnosis

Primarily for the purpose of identification of eye problems, image processing has been used in clinical settings for a considerable amount of time. Patients who suffer from eye illnesses will see changes in the picture arrangement and structure of the different sections of the eye. For the purpose of photographic and researching various elements of retinal architecture and operate, numerous imaging devices such as the The funds camera, Scanning Laser Ophthalmoscope (SLO), and Optical Coherence Tomography (OCT) make it feasible to see and investigate these characteristics. These colour fundus photos are used to identify DR. This picture is used to separate the layers of retinal structures, such as the vessels that carry blood then the optic disc, and then confirmed by images taken from the ground truth. An efficient division of the optic disc and circulation vessels, correspondingly, was achieved as a consequence of the technique of segmentation. A fuzzy logic-based algorithm was developed for the purpose of classifying non-proliferative diabetes-related retinopathy as usual, mild, severe, moderate, and extremely advanced. This categorising was established on the number of disease-related signs, which included haemorrhages, tiny arteries, and discharge in fundus pictures. Performing better than competitors like Inception, ResNet101, and ResNet50, the suggested fuzzy inference system attained a remarkable 97% accuracy. The results show that the fuzzy logic method outperforms the traditional deep neural network models when it comes to retinopathy with diabetes stage classification.