Purchase this article with an account.
Charles Reisman, Ying Dong, Qi Yang, Wei-Chieh Huang, Tsutomu Kikawa; Comparison of Bruch's membrane opening detection between DRI OCT Triton and Zeiss Cirrus in a myopic cohort. Invest. Ophthalmol. Vis. Sci. 2018;59(9):2093.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To present the disc detection results on a study comparing Deep Range Imaging Optical Coherence Tomography (DRI OCT) Triton and Zeiss Cirrus in a cohort of myopic subjects.
52 glaucomatous eyes were imaged in three clinical sites using DRI OCT Triton (Topcon, Tokyo, Japan) and Zeiss Cirrus 4000/5000 (Carl Zeiss Meditec Inc., Dublin, USA). Each eye was imaged with three or more Triton wide fixation 12x9mm2 vertical (V) scan volumes per visit, and three or more Cirrus 3D disc cube horizontal (H) scans per visit, two visits in total. Exclusion criteria included blinks, eye motion, large floaters, clipping, feature centration, and segmentation errors. 382 Triton scans and 301 Cirrus scans from the same eyes were accepted for the analysis. Scleral ring disc margin points and corresponding disc area measurements were detected automatically using officially released Fastmap software v10.50 in DRI OCT Triton. Triton Bruch’s membrane opening (BMO) based disc margin points and disc area measurements were automatically detected using software currently in development (V184.108.40.206). Cirrus BMO detection was performed using commercially available software v220.127.116.112. The two BMO based disc areas and one scleral ring based disc area were compared based on the first accepted scans. Deming regression was used to examine agreement. In addition, a visual check on individual B-scans was performed to determine the correctness of BMO point detection.
The average disc area was 2.391±0.087 mm2 for Triton BMO based disc detection, 1.916 ± 0.053 mm2 for Cirrus BMO based disc detection, and 2.018 ± 0.112 mm2 for Triton scleral ring based disc detection. The agreement plots between Triton BMO and Cirrus BMO, Triton scleral ring and Cirrus BMO are shown in Fig. 1. The visual check reveals that Cirrus’ BMO disc detection excluded border tissue when it was present, thereby corresponding to scleral ring points.
A significantly better agreement between Triton scleral ring and Cirrus BMO disc area measurements was observed compared to the agreement between Triton BMO and Cirrus BMO disc area measurements. The agreement results and visual comparison indicates that Cirrus may detect scleral ring points rather than BMO points in a myopic cohort.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.
Fig.1 Agreement plots between Triton BMO and Cirrus BMO, and Triton scleral ring and Cirrus BMO based disc area.
This PDF is available to Subscribers Only