Machine Learning Techniques for Ophthalmic Data Processing: A Review.

Area of Science:

• Ophthalmology
• Medical Imaging
• Artificial Intelligence

Background:

• Machine learning (ML) and deep learning (DL) are increasingly vital in analyzing medical data.
• Ophthalmic disease diagnosis relies heavily on interpreting complex medical images.
• Accurate and early detection of diseases like diabetic retinopathy, age-related macular degeneration, and glaucoma is crucial.

Purpose of the Study:

• To comprehensively review ML approaches for diagnosing three major ophthalmic diseases over the past four years.
• To summarize public datasets, challenges, and current methodologies in ophthalmic disease detection, grading, and segmentation.
• To explore the application of ML in retinal vessel segmentation, retinal layer segmentation, and fluid segmentation.

Main Methods:

• Systematic review of over 60 publications and 25 public datasets.
• Analysis of ML techniques applied to color fundus imaging and optical coherence tomography.
• Inclusion of ML methods using eye measurements and visual field data for glaucoma detection.

Main Results:

• Detailed summary of ML applications for diabetic retinopathy, age-related macular degeneration, and glaucoma diagnosis.
• Overview of segmentation techniques for retinal vessels, layers, and fluid.
• Identification of common challenges and available datasets for each disease.

Conclusions:

• ML shows significant promise for advancing ophthalmic disease diagnosis and analysis.
• Further research is needed to address limitations and integrate these techniques into clinical practice.
• The review provides a valuable resource for researchers and clinicians in the field of ophthalmic AI.