Abstract
Purpose :
To compare foveal avascular zone (FAZ) measurements of different three commercially available optical coherence tomography-angiography (OCT-A) machines.
Methods :
In a cross-sectional prospective non-interventional study of 27 healthy volunteers, foveal-centered 3×3mm OCT-A scans obtained by three machines: DRI-OCT Triton; TOPCON (Triton), RS3000 Advance; NIDEK (RS3000) and CIRRUS HD-OCT model 5000; ZEISS (CIRRUS). FAZ in the superficial and deep plexus layers were measured by two masked graders. The intrarater, interrater, and intermachine agreements were assessed.
Results :
In each machine, intrarater correlation coefficient was higher than 0.964 (P<0.001). Interrater correlation was also higher than 0.987 (P<0.001). For superficial FAZ, intermachine correlation coefficient (95% confidence interval) was 0.920 (0.803-0.965) for Triton versus RS3000, 0.899 (0.575-0.965) for RS3000 versus CIRRUS and 0.963 (0.913-0.983) for CIRRUS versus Triton (P<0.001). For deep FAZ, intermachine correlation coefficient was 0.813 (0.633-0.910) for Triton versus RS3000, 0.694 (0.369-0.857) for RS3000 versus CIRRUS and 0.679 (0.153-0.872) for CIRRUS versus Triton (P<0.001). The average of superficial FAZ was 0.264 ± 0.071 mm2 with Triton, 0.278 ± 0.072 mm2 with RS3000 and 0.257 ± 0.066 mm2 with CIRRUS. There was a significant difference between any two of three machines. Brand-Altman analysis revealed the existence of systemic error between machines. In deep FAZ, the average FAZ was 0.617 ± 0.175 mm2 with Triton, 0.646 ± 0.178 mm2 with RS3000 and 0.719 ± 0.175 mm2 with CIRRUS. There was a significant difference between RS3000 versus CIRRUS or CIRRUS versus Triton.
Conclusions :
FAZ measured with three commercially available OCT-A machines showed excellent reproducibility and repeatability in healthy eyes, indicating that they are clinically interchangeable to a certain extent. However, the absolute value of FAZ was significantly different between machines. This information should be remembered in research or mass screening using FAZ.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.