Zeiss CIRRUS HD-OCT 500, User Manual

Page: 317 / 352

CIRRUS HD-OCT User Manual 2660021169012 Rev. A 2017-12
Appendix
B-9
Study 2: Repeatability and Reproducibility of Illumination Area Measurements Under the
Retinal Pigment Epithelium
The objective of this study was to determine the repeatability and reproducibility of the
CIRRUS HD-OCT measurement of illumination areas under the retinal pigment epithelium
(RPE). For Phase 1, the inter–device analysis, 49 eyes of 37 subjects were qualified for
inclusion. A single operator performed all scans during Phase 1 of the protocol. For Phase
2, the inter–operator phase, 53 eyes of 39 subjects were qualified for inclusion into the
analysis of scans from a single CIRRUS HD-OCT taken by three operators. Subjects in each
phase of the study underwent the following series of CIRRUS HD-OCT scans:
Phase 1: Inter–Device Variability
• Three acceptable Macular Cube 200x200 scans on one eye from three CIRRUS
HD-OCT instruments for a total of 9 scans
• Three acceptable Macular Cube 512x128 scans on the fellow eye from three CIRRUS
HD-OCT instruments for a total of 9 scans
Phase 2: Inter–Operator Variability
• Three acceptable Macular Cube 200x200 scans on one eye from one CIRRUS HD-OCT
instrument by three operators for a total of 9 scans
• Three acceptable Macular Cube 512x128 scans on the fellow eye from the same
CIRRUS HD-OCT instrument by three operators for a total of 9 scans
The repeatability standard deviation (SD) is the square root of the random variance
component. The reproducibility SD is the square root of the sum of all contributions to
variance except subject variance. The repeatability limit is 2.8x repeatability SD. The
reproducibility limit is 2.8x the reproducibility SD. The larger random error variation from
the two study phases was selected as the random error variation for the corresponding
endpoint (and scan type) and was used to calculate the repeatability. The reproducibility
includes variation due to random error, operator, device, interaction between subject and
device, and interaction between subject and operator.
Table B-8 shows the repeatability and the reproducibility SD and limits of the sub–RPE
illumination area measurements and the closest distance to the fovea determined by the
automated algorithm for both 200x200 and 512x128 scans. For the area measurements,
the repeatability limit of the 200x200 vs. the 512x128 scans were very similar (2.4885 and
2.4313 mm
2
respectively). The reproducibility limit of the 200x200 scan was smaller
compared to that of the 512x128 scans (2.6460 and 2.8889 mm
2
respectively). The
repeatability and reproducibility limits will affect the ability to determine when
measurements have changed due to a change in pathology as opposed to random
variability.