This study focused on classifying large sets of Optical Coherence Tomography (OCT) retinal fundus images into categories indicative of either healthy retinas or those affected by diabetic retinopathy. To achieve accurate classification, this study employed advanced feature extraction methods, specifically the Gray Level Run Length Matrix (GLRLM), the Gray Level Co-Occurrence Matrix (GLCM), and Gray Level Histogram Features (GLHFs). These techniques are crucial for capturing detailed textures and patterns within the retinal images, which are instrumental in distinguishing between healthy and disease-affected tissues. A total of 301 color OCT retinal fundus images were analyzed in this research. These images were sourced from both healthy individuals and those diagnosed with diabetic retinopathy, providing a comprehensive dataset for evaluation. To enhance the quality of the images and improve the accuracy of the feature extraction, a fourth-order Partial Differential Equation (PDE) filter was applied during the image pre-processing phase. This filtering step aimed to reduce noise and enhance the structural features in the images. The primary objective of this study was to identify the most effective feature extraction technique for differentiating between healthy and diabetic retinopathy-affected retinas. By comparing the performance of the GLRLM, the GLCM, and GLHFs, this study sought to determine which method offers the most reliable results in retinal disease classification, thus contributing to better diagnostic tools and methodologies in ophthalmology.