Localizing Optic Disc in Retinal Image Automatically with Entropy Based Algorithm

Area of Science:

• Ophthalmology
• Medical Imaging
• Computer Vision

Background:

• Continuous examination of retinal images is crucial for detecting eye diseases.
• Automated and accurate computer-aided scanning of retinal images is needed for efficient diagnosis.
• Optical disc localization is a fundamental step in analyzing retinal image elements.

Purpose of the Study:

• To propose a novel algorithm for automated optical disc localization in retinal images.
• To enhance the accuracy and efficiency of retinal image analysis for eye health assessment.
• To develop a computer-aided tool for identifying key retinal structures.

Main Methods:

• The proposed algorithm utilizes the higher entropy value of the optical disc area for identification.
• A sliding window technique is employed to systematically analyze image patches.
• Recursive steps are used to test the entropy of various image patches.

Main Results:

• The algorithm achieved good accuracy in localizing the optical disc.
• The method demonstrated efficiency with trivial computation time.
• Practical results were validated using common retinal image datasets.

Conclusions:

• The developed algorithm offers an accurate and efficient solution for optical disc localization.
• This computer-aided approach can significantly contribute to automated eye health monitoring.
• The entropy-based method provides a promising direction for retinal image analysis.