| Yazarlar |
Doç. Dr. Kemal AKYOL
Kastamonu Üniversitesi, Türkiye |
| Özet |
| Diabetic retinopathy is a disease that causes blindness resulting from damages that emerge in the retina depending on the diabetes mellitus. There are two stages of the disease including the non-proliferative and proliferative. Eyesight loss is blocked by means of early detection and diagnosis of non-proliferative DR findings. In this study, we designed a decision support system for automatic detection of hard exudates which are early stage DR lesions. This system consists of keypoint extraction, feature extraction, visual dictionary and classifying stages. We tested the performance of the system, which we carried out based on system learning and analysis of new retinal images, on the public DIARETDB1 retinal image dataset. Experimental results obtained with Artificial Neural Networks, Random Forest ve Decision Tree algorithms showed us that machine learning technique suggested by us is successful. |
| Anahtar Kelimeler |
| Makale Türü | Özgün Makale |
| Makale Alt Türü | Uluslararası alan indekslerindeki dergilerde yayımlanan tam makale |
| Dergi Adı | Mugla Journal of Science and Technology |
| Dergi Tarandığı Indeksler | |
| Makale Dili | İngilizce |
| Basım Tarihi | 01-2016 |
| Cilt No | 2 |
| Sayı | 1 |
| Sayfalar | 1 / 6 |
