期刊名称：International Journal of Computer Trends and Technology
电子版ISSN：2231-2803
出版年度：2018
卷号：61
期号：1
DOI：10.14445/22312803/IJCTT-V61P109
出版社：Seventh Sense Research Group
摘要：Diabetic retinopathy (DR) is a retinal disease that is affected by diabetes on the eyes. The main risk of the disease can lead to blindness. Control of the blood glucose levels in the blood system depends on glucose and insulin interaction. Medical image diagnosis plays a major role of research for Health care purposes. DR can be occurred because enough rate of insulin in the body is not secreted properly by the pancreas. If a person has diabetes for 20 years or more, he or she has the more probability to suffer diabetic retinopathy. DR usually shows no symptoms or vision problems at early stage of the disease. However, it can lead blindness eventually. The earliest clinical sign of DR is the detection of microaneurysms (MAs). They are formed due to the leakage of blood from capillary. MAs are small, red dots and spread on the superficial retinal layers. Diabetics risk likely to be developed from a person’s daily life style activity such as his/her eating habits, sleeping habits, physical activity and so on. In the proposed work, screening algorithm and KNN algorithm is used for Screening of DR to prevent blindness. In this algorithm lesions of eye including blood vessel exudates and micro aneurysms are detected using morphological operations. The retina images from standard neural network disease diabetic retinopathy database and local database are used as inputs of proposed work. Thus DR can be detected using the proposed work and it will be useful for further treatment.
Loading...
