<![CDATA[Retinal diseases such as choroidal neovascularization (CNV), diabetic macular edema (DME), and drusen are the leading causes of vision impairment and require early detection through optical coherence tomography (OCT) imaging. The diagnostic process, which is performed manually by ophthalmologists, is relatively time-consuming and may lead to delays in treatment. This study aims to develop a Convolutional Neural Network (CNN)-based retinal condition detection application integrated with a desktop application to assist in the automatic analysis of OCT images. The data used comes from the Kermany OCT Dataset, which consists of 30,000 retinal images divided into four categories: CNV, DME, drusen, and normal. The research stages include image preprocessing, such as resizing to 224×224 pixels, normalization, contrast enhancement using Contrast Limited Adaptive Histogram Equalization (CLAHE), and data augmentation. The CNN model was developed using Python with the TensorFlow and Keras libraries to extract image features and classify retinal conditions. Test results show that the CNN model achieved an accuracy rate of 99.73% in classifying retinal OCT images. The trained model was then integrated into a Java-based desktop application so it can be used as a diagnostic support system to facilitate faster and more consistent retinal image analysis. The results of the study indicate that the CNN method is effective for classifying retinal diseases and has the potential to support the early detection of retinal diseases based on OCT images.]]>
Copyrights © 2026
