Abstract
Purpose :
To compare retinal blood flow velocity measurements using the XyCAM RI across multiple operators to assess the repeatability and reproducibility of the imager.
Methods :
A total of twenty patients were imaged using the XyCAM RI. Group 1 contained ten patients that were imaged by a single operator, with four imaging sessions centered around the optic nerve head (ONH) and four imaging sessions centered around the macula. Groups 2 and 3 contained five patients each that were imaged by three operators, with each operator acquiring four imaging sessions centered around the ONH in Group 2 and four imaging sessions centered around the macula in Group 3. All imaging sessions were analyzed using the XyCAM RI Software. For ONH centered sessions, the following regions were analyzed: optic disc, major artery, major vein, branch point, and a branched vessel (Figure 1). For macula centered sessions, the following regions were analyzed: fovea, superior temporal region, superior nasal region, inferior nasal region, and inferior temporal region (Figure 1). For Groups 2 and 3, the same regions were analyzed by the three different operators. Each imaging session was analyzed by operator selected cardiac cycles, characterized by minimal movement and imaging artifacts. The intra-session coefficient of variation (CV) was calculated across the cardiac cycles while the inter-session CV was calculated across the imaging sessions. CV was additionally calculated for subjects in Groups 2 and 3 across the three operators.
Results :
The average intra-session CV was 2.88 ± 1.05% in the ONH region and 2.49 ± 0.67% in the macular region across all operators. The average inter-session CV was 6.45 ± 3.78% in ONH centered sessions and 6.49 ± 1.34% in macula centered sessions across all operators. Across the three operators, the average CV was 6.92 ± 3.38% in ONH centered sessions and 5.94 ± 2.33% in macula centered sessions. Figure 2 displays the CVs for each individual region and operator.
Conclusions :
The XyCAM RI is able to produce repeatable retinal blood flow velocity measurements across multiple cardiac cycles within a single imaging session as well as across multiple imaging sessions. There is also a high reproducibility of measurements as observed by calculating the variation measurements acquired by three different operators.
This is a 2020 ARVO Annual Meeting abstract.