Abstract
Purpose :
To investigate the reliability of quantitative measurements of the choriocapillaris (CC) using spectral domain optical coherence tomography angiography (SD-OCTA) in healthy individuals.
Methods :
In this IRB-approved prospective study, one eye of healthy normal subjects was imaged with a full-spectrum amplitude-based probabilistic SD-OCTA system (Spectralis OCT2; Heidelberg Engineering, Heidelberg, Germany). A standardized acquisition protocol was employed with OCTA scans centered on the fovea, covering a 15° x 15° area (768 A-scan x 768 B-scans). Scans were repeated three times. En face OCTA slabs were extracted for multiple layers including the CC. The three repeat scans were aligned by first performing registration on the superficial capillary plexus images, and then applying this transformation to the CC images. The 3 registered CC images were then averaged. Single (three separate) and averaged CC images were then exported to ImageJ and binarized to allow computation of the CC flow void/deficit area which was expressed as a percentage of the total scan area. The CC flow deficit was compared between the repeat single (unaveraged) scans to assess repeatability, and also compared between single and averaged scans.
Results :
This study included 22 normal eyes of 22 subjects (14 male, 8 female) with a mean age±standard deviation of 34.0±6.9 years (range: 22-55 years). For this cohort, the mean total flow deficit for the 3 single, unaveraged CC images were 50.7%, 51.8%, and 53.6%, respectively. The intraclass correlation coefficient among these single images was 0.932. Compared with these single unaveraged images, the averaged images showed a significantly smaller total flow deficit (25.0% [95% CI, 17.9%-30.6%] vs 52.0% [95% CI, 43.0%-61.0%], P<0.0001). The sinusoidal meshwork architecture of the CC was more clearly evident on the averaged CC images.
Conclusions :
In healthy individuals, measurements of the CC using full-spectrum SD-OCTA demonstrated good repeatability even when using single, unaveraged images. Averaging of multiple images improves the CC appearance, but also produces a lower flow deficit measurement.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.