Assessment of retinal microvasculature in supine vs upright positioning using OCT-Angiography

Abstract

Purpose : Recent development of portable OCT-A devices has increased the scenarios in which OCTA can be utilized and the populations amenable to screening. It is pertinent to gain an understanding of the differential effects of upright positioning used in clinic vs supine positioning used in the OR on OCTA images as well as validate the equivalence of portable devices. We performed a cross sectional clinical observational study to determine if there are position and machine dependent differences in OCTA images in normative adult patients and to better understand the dynamics of retinal blood flow with position.

Methods : We obtained 10x10 scan angle images of right eye maculae of 41 healthy adult participants at 90° and 0° using OCTA integrated Spectralis SD-OCT tabletop and investigational portable Flex system (Fig. 1). A customized MATLAB algorithm was used to calculate FAZ, VAD, and VLD (Fig. 2). Wilcoxon signed rank tests were run to determine how FAZ, VAD, and VLD varied between imaging scenarios and assess differential contribution of age.

Results : From Oct. to Nov. 2019, 41 subjects were imaged with the tabletop system at 90° and with the Flex system at 90° and 0°; only images of acceptable quality were used for analysis (n=39, mean age 41.6 ± 14.3 yrs, range 20-68 yrs). FAZ, VAD, and VLD were measured for 39 eyes. A statistically significant difference was found for FAZ and SVC VLD between the 2 imaging systems in the upright position (p=.026, p=.036). Significant positional differences were only found in the younger half of the overall group with higher SVC VAD and VLD in the supine position (<42 y.o; p=.04, p=.011).

Conclusions : There appear to be small differences at baseline between the tabletop and Flex OCTA systems in the upright position. Positional differences seen in VAD and VLD in younger vs. older subjects might suggest more robust changes in vasoconstriction with lower age. Further exploration is necessary to better inform interpretation of images obtained in the clinic compared to the OR, especially in younger populations.

This is a 2020 ARVO Annual Meeting abstract.

 

Fig 1. SVC in healthy adult with tabletop and investigational Flex: 1A) Upright tabletop, 1B) Upright Flex, and 1C) Supine Flex Positioning. 1D-F) Corresponding SVC scans.

 

Fig 2. Custom MATLAB processing of OCTA images of DVC of healthy adult: 2A) Original Image, 2B) Binarized, 2C) Skeletonized, 2D) Filtered, 2E) Traced FAZ from filtered, 2F) Cropped FAZ.