KLASIFIKASI RETINOPATI DIABETIK MENGGUNAKAN JARINGAN SARAF TIRUAN
| Main Author: | Pangestu, Iwan |
|---|---|
| Format: | Thesis NonPeerReviewed Book |
| Bahasa: | ind |
| Terbitan: |
, 2015
|
| Subjects: | |
| Online Access: |
https://eprints.untirta.ac.id/9380/1/Klasifikasi%20Retinopati%20Diabetik%20Menggunakan%20Jaringan%20Saraf%20Tiruan.pdf https://eprints.untirta.ac.id/9380/ https://ft.untirta.ac.id |
Daftar Isi:
- Diabetic Retinopathy is a disorder of the blood vessels in retinal arising as a result of the chronic complications of diabetes mellitus. A medical examination of patients with diabetic retinopathy disease carried by direct observation in the retinal image using a fundus camera, and require additional time to analyze the image of the patient by the ophthalmologist. The severity of diabetic retinopathy is divided into three classes, namely: Non diabetic retinopathy (NDR), Non- Proliferative Diabetic Retinopathy (NPDR), and Proliferative Diabetic Retinopathy (PDR). This study aims to determine the optimal parameters of the neural network for the classification of the severity of diabetic retinopathy disease. Statistical Characteristics of patients retinal image obtained by using extraction process with morphology extraction method. Statistical characteristic are then trained using neural network with 4 kind of backpropagation learning method to obtain optimal results in the classification. The optimal results of the classification of the severity disease of diabetic retinopathy using the learning algorithm Levenberg-Marquardt backpropagation 2 hidden layer with sensitivity on NDR class of 100%, spesificity on NDR class of 91.67%, the sensitivity on NPDR class of 81.82%, spesificity on NPDR class of 95.45%, the sensitivity on PDR class of 92.30%, spesificity on PDR class of 100% and accuracy of 90,91%. Keywords: diabetic retinopathy, Levenberg-Marquardt Backpropagation, neural networks, backpropagation.