OCT feature analysis guided artery-vein differentiation in OCTA.

Area of Science:

• Ophthalmology
• Medical Imaging
• Biomedical Engineering

Background:

• Differential artery-vein analysis in retinal imaging enhances disease detection sensitivity.
• Current clinical optical coherence tomography angiography (OCTA) systems lack automated artery-vein differentiation.
• Accurate artery-vein identification is crucial for classifying retinal vascular diseases.

Purpose of the Study:

• To investigate the feasibility of using optical coherence tomography (OCT) intensity features for artery-vein differentiation.
• To develop a method for generating an OCT-based artery-vein map to guide OCTA analysis.
• To improve the diagnostic capabilities of OCTA by enabling direct artery-vein classification.

Main Methods:

• Analysis of four OCT intensity profile features: vessel width to central reflex ratio, average maximum profile brightness, average median profile intensity, and boundary optical density.
• Classification of OCT artery-vein source nodes using these intensity features.
• Application of a blood vessel tracking algorithm to automatically generate an OCT artery-vein map.
• Direct application of the OCT artery-vein map to guide artery-vein differentiation in OCTA.

Main Results:

• Identified specific OCT intensity features capable of distinguishing between arterial and venous vessels.
• Successfully generated an automated OCT artery-vein map.
• Demonstrated that the OCT artery-vein map can directly guide artery-vein differentiation in OCTA.
• Validated the feasibility of using OCT data for artery-vein classification in OCTA.

Conclusions:

• OCT intensity feature analysis is a viable method for differentiating retinal arteries and veins.
• The developed OCT artery-vein mapping technique can enhance OCTA's sensitivity for retinal disease detection and classification.
• This approach offers a pathway to integrate artery-vein differentiation into clinical OCTA workflows without requiring new hardware.