Non-Invasive Retinal Pathology Assessment Using Haralick-Based Vascular Texture and Global Fundus Color Distribution...

Area of Science:

• Ophthalmology
• Medical Imaging
• Biomedical Engineering

Background:

• Diabetic retinopathy (DR) and glaucoma are leading causes of vision loss.
• Early detection and classification of these retinal diseases are crucial for effective treatment.
• Current diagnostic methods may be invasive or lack comprehensive analysis of retinal features.

Purpose of the Study:

• To develop and validate a dual-framework approach for distinguishing healthy retinas from those with DR and glaucoma.
• To analyze vascular texture and global color distribution for disease-specific biomarkers.
• To assess the potential of these features for non-invasive early disease detection.

Main Methods:

• Retinal fundus images were analyzed using vascular texture analysis (Haralick features from Gray-Level Co-occurrence Matrix) and global color distribution analysis (RGB histograms).
• Texture features (energy, contrast, correlation, entropy) were extracted after retinal vasculature segmentation.
• Global color distribution was analyzed without segmentation, with Mahalanobis distance used to quantify group separability.

Main Results:

• Significant differences in texture features were observed between healthy and pathological retinas, with diseased retinas showing lower complexity.
• Pathological retinas exhibited higher uniformity in texture, correlating with vascular thinning and structural changes.
• Significant chromatic shifts were detected in the green and blue channels of pathological retinas, indicating altered tissue properties.

Conclusions:

• Vascular texture analysis provides valuable insights into structural changes associated with DR and glaucoma.
• Global color distribution analysis reveals detectable chromatic shifts in diseased retinas.
• Both vascular texture and color features show potential as non-invasive biomarkers for early retinal disease detection and classification.