Intra- and Inter-Subject Variability of Retinal Oximetry on Healthy Eyes Using Visible-Light OCT

Abstract

Purpose : To evaluate intra- and inter-subject variability of the oxygen saturation (sO₂) measurements in major retinal blood vessels using visible-light optical coherence tomography (vis-OCT).

Methods : Four eyes of 4 healthy volunteers were enrolled. For each eye, one pair of retinal major artery and vein in superior and inferior regions near the optic nerve head (Fig.) were repeatedly scanned four times during the same session with our prototype vis-OCT (8192 A-scans over 1mm², sampling interval = 0.12 µm). Spectroscopic analysis was done on the raster scans using short-time Fourier Transform (STFT). For reliable sO₂ estimation, wavelength-dependent OCT amplitude from same depth location across multiple A-lines were averaged. Coefficient of variations (CVs) were calculated to assess sO₂ measurement repeatability for each subject/location. Intraclass correlations (ICCs) were calculated to assess inter-subject variability on sO₂ difference between the paired artery and vein.

Results : Subjects demographics and average sO₂ measurements are shown in Table 1. Intra-subject measurement repeatability was very high (CVs ranging from 0.21 to 1.76%, Table 2). The retinal arteries showed consistently higher sO₂ than the paired veins regardless the location. The sO₂ differences between the paired artery and vein showed small variability even among subjects (Table 3).

Conclusions : The sO₂ measurements showed high intra-subject repeatability and low inter-subject measurement variability. Vis-OCT technology may be a useful and reproducible clinical tool for performing retinal oximetry in vivo.

This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.

 

Figure. Sampling locations of retinal oximetry. A pair of artery (A) and vein (V) in superior(sup) and inferior(inf) regions near the optic nerve head are sampled to measure sO₂ (highlighted dashed lines).