Compensation for Reflectance Variation in Vessel Density Quantification by Optical Coherence Tomography Angiography

Area of Science:

• Ophthalmology
• Medical Imaging
• Biomedical Engineering

Background:

• Optical coherence tomography angiography (OCTA) is crucial for quantifying retinal and optic nerve head (ONH) vasculature.
• Reflectance variations in OCTA scans can introduce noise and affect the accuracy of vessel density measurements.
• Accurate vessel density quantification is essential for diagnosing and monitoring various ocular diseases.

Purpose of the Study:

• To develop and validate a method for compensating reflectance variations in OCTA.
• To improve the reliability and accuracy of vessel density quantification using OCTA.
• To reduce inter-individual variability in OCTA-derived vessel density measurements.

Main Methods:

• Utilized a split-spectrum amplitude-decorrelation angiography (SSADA) algorithm on a spectral-domain OCT system for macular and ONH scans.
• Developed a reflectance-adjusted decorrelation threshold equation based on the linear relationship between OCT reflectance and OCTA noise.
• Evaluated background OCTA noise in the foveal avascular zone (FAZ) and calculated vessel density from binarized en face OCTA images.

Main Results:

• A reflectance-adjusted threshold effectively filtered out 97.5% of background OCTA noise.
• The correlation between vessel density and reflectance signal strength index (SSI) was eliminated using the adjusted threshold.
• Reflectance compensation reduced population variation from 8.5% to 4.8% in the macula and improved within-visit repeatability from 4.4% to 1.8%.

Conclusions:

• Compensating for reflectance variation significantly enhances the reliability of OCTA vessel density quantification.
• The developed reflectance-adjusted thresholding method offers improved accuracy and reduced variability in OCTA analysis.
• This technique holds promise for more precise assessment of ocular vasculature in clinical and research settings.