Review on Microaneurysms Detection Using Image Processing

It has been very challenging to forecast the residing microaneurysms in fundus images and recognize diabetic retinopathy in its early stages for a very long time. Long-term high blood glucose levels cause diabetic retinopathy (DR), which has irreversible microvascular consequences of loss of vision. Since microaneurysms (MAs) are thought to be the first lesions in diabetic retinopathy, their detection is crucial for the diagnosis of the condition. Recently, a variety of machine-learning methods for MA identification have been developed. Typically, these candidates' features are extracted after MA candidates have been discovered, the necessity for early detection of the most typical signs is stronger than ever due to the high incidence of diabetes mellitus and its effects. The presence of microaneurysms in the fundus pictures is one of the initial symptoms. The identification of red lesions in colored retinal imaging represents a indispensable first step in preventing the progression of sight impairment and visual loss brought on by DR. This makes their detection necessary for DR mass screening. This article presents a brand-new, comprehensive method that considers both pathological risk factors and the turnover of MAs when diagnosing the development of DR. In particular, one method obtains MAs turnover by following the conventional image analysis-based road map.