Optic Disk Edema Rough Classification: A Linear SVM Approach with Visual Features

Optic disk edema (ODE), a notable retinal anomaly, is crucial to prescreen and grade for severity to prevent further complications. In this study, 44 visual optic disk features derived from texture, color, vessels, size, and the intensity of the inner and outer area of the optic disk were extracted. Minimum redundancy maximum relevance (mRMR) was combined with support vector machine (SVM) techniques to systematically select the top essential features for optimal performance. The proposed method achieved an accuracy of 87.91%, a specificity of 94.50%, and a sensitivity of 82.19%. This performance outperformed existing approaches, demonstrating the effectiveness of the proposed method in differentiating between healthy optic disks and ODE, as well as in grading the severity of ODE.