Algorithms for red lesion detection in Diabetic Retinopathy: A review

Area of Science:

• Ophthalmology
• Medical Imaging
• Computer-Aided Diagnosis

Background:

• Diabetic Retinopathy (DR) is a complication of diabetes affecting vision.
• Early stages of DR are often asymptomatic, leading to delayed diagnosis and irreversible vision loss.
• Microaneurysms (MA) are early indicators of DR, making their detection critical for timely intervention.

Purpose of the Study:

• To review and systematize state-of-the-art automated methods for analyzing microaneurysms and hemorrhages in medical images.
• To provide a qualitative and quantitative comparison of existing literature on DR analysis.
• To identify limitations in current algorithms and guide future research in computer-aided DR diagnosis.

Main Methods:

• Comprehensive literature review of automated computer-aided diagnosis techniques for DR.
• Analysis of methods focusing on the detection of microaneurysms and hemorrhages.
• Qualitative and quantitative comparison of identified algorithms and their performance.

Main Results:

• Overview of various state-of-the-art automated methods for microaneurysm and hemorrhage detection.
• Comparative analysis highlighting the strengths and weaknesses of existing approaches.
• Identification of research gaps and limitations in current automated DR analysis.

Conclusions:

• Automated analysis of medical images shows promise for early DR detection and management.
• Systematization of current methods is essential for guiding researchers and improving diagnostic accuracy.
• Further research is needed to overcome limitations in automated microaneurysm and hemorrhage analysis for DR screening.