Automatic screening and classification of diabetic retinopathy and maculopathy using fuzzy image processing

Area of Science:

• Ophthalmology
• Medical Imaging
• Computer Science

Background:

• Diabetic retinopathy and maculopathy require regular screening to prevent visual impairment.
• Current digital retinal imaging screening methods lack effectiveness, robustness, and cost-efficiency.

Purpose of the Study:

• To develop a novel automatic detection system for diabetic retinopathy and maculopathy using fuzzy image processing.
• To introduce a new technique for macula region localization to enhance maculopathy detection.
• To present a new online dataset for eye fundus images to aid diabetic retinopathy research.

Main Methods:

• Employing fuzzy image processing techniques combined with Circular Hough Transform.
• Implementing a four-part system: image acquisition, preprocessing with retinal structure localization, feature extraction, and classification.
• Developing a novel macula region localization technique for improved maculopathy detection.

Main Results:

• The proposed system successfully integrates fuzzy logic, Hough Transform, and feature extraction for automated detection.
• A novel method for macula localization was developed and implemented.
• A new online dataset of eye fundus images was created, detailing its collection and expert diagnosis process.

Conclusions:

• The developed fuzzy image processing system offers a promising approach for automatic diabetic retinopathy and maculopathy detection.
• The novel macula localization technique enhances the system's capability in identifying maculopathy.
• The new online dataset provides a valuable resource for advancing diabetic retinopathy research and screening development.