Advanced Feature Extraction from Retinal Color Fundus Images for Accurate Macular Disease Diagnosis

Image processing has been used in the field of biomedicine for quite some time, particularly for the purpose of diagnosing eye problems. When an individual has an eye ailment, the visual structure and smoothness of all of the layers of their retinas will be altered. A number of imaging devices, such as the Fundus their cameras, the Scanner Laser Ophthalmoscope (SLO), and the Optical Coherence Tomography (OCT), have helped make it feasible to image and investigate several facets of the structure and functionality of the retina. Diabetic retinopathy (DR) is another illness that is prevalent among those who have glucose mellitus. In addition to haemorrhages, tiny arteries, and an abundance of both hard as well as soft exudates, it is characterised by arteries that are leaking and have an uneven form. There is a possibility that the blood arteries may burst, causing blood to spill into the retina. Additionally, it causes fluid to accumulate close to the macula, which results in a reduction in vision. Proliferative retinopathy due to diabetes is a more serious kind of the illness, whereby the veins that have been destroyed will become clogged, which will then lead to the abnormal development of fresh vessels that leak. When it comes to these disorders, early identification and management are very important, since they might otherwise result in blindness. Compared to competing approaches like VGG16, GoogleNet, Inception V3, and CSR, the suggested approach of segmenting optic discs and blood vessels segmentation has a 99% success rate. As a result, the suggested method outperforms the alternatives when it comes to processing retinal pictures properly.