Detection of Diabetic Retinopathy Using Discrete Wavelet-Based Center-Symmetric Local Binary Pattern and Statistical...

Area of Science:

• Ophthalmology
• Medical Imaging
• Computer Science

Background:

• Diabetic retinopathy (DR) is a leading cause of blindness, necessitating early detection.
• Computer-aided diagnosis (CAD) systems offer automated analysis of retinal images for timely DR intervention.

Purpose of the Study:

• To introduce a novel CAD system for enhanced early diabetic retinopathy detection.
• To evaluate the efficacy of multi-resolution and global feature analysis for retinal image classification.

Main Methods:

• Retinal images were preprocessed using median filter, CLAHE, and unsharp filters to improve quality.
• Discrete Wavelet Transform (DWT) was used for multi-resolution image decomposition.
• Center Symmetric Local Binary Pattern (CSLBP) and statistical features were extracted from decomposed images.

Main Results:

• The proposed CAD system demonstrated superior performance compared to existing methods on a benchmark dataset.
• Wavelet-based CSLBP features effectively captured fine and coarse details of retinal images.
• Combining wavelet-based CSLBP features with statistical features significantly improved detection performance.

Conclusions:

• The novel CAD system shows significant promise for accurate and early diabetic retinopathy detection.
• Multi-resolution analysis combined with global feature extraction provides a comprehensive approach for retinal image analysis.
• The integration of CSLBP and statistical features offers a robust strategy for improving CAD system performance in DR detection.