Purchase this article with an account.
Ying Dong, Qi Yang, Wei Chieh Huang, Charles A Reisman; Comparison of Bruch's membrane opening and scleral ring disc margin detection methods using DRI OCT Triton scan data. Invest. Ophthalmol. Vis. Sci. 2017;58(8):4011.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To present disc detection results on a study comparing Bruch’s membrane opening (BMO) based method and the scleral ring based method using the scan data from Deep Range Imaging Optical Coherence Tomography (DRI OCT) Triton, a one-micron wavelength swept-source OCT device.
21 normal eyes and 11 glaucoma eyes were imaged in a clinical site using two DRI OCT Triton machines (Topcon, Tokyo, Japan). Each eye was imaged with three or more scans for each scan mode: three-dimensional optic disc 6x6mm2 horizontal (H) scans, wide fixation 12x9mm2 horizontal (H) scans, and wide fixation 12x9mm2 vertical (V) scans. The data had exclusions performed based on clinical criteria, which included checks for blinks, eye motion, clipping, feature centration, segmentation errors, local weak signal, and image quality. 211 disc 6x6H, 211 wide 12x9H, and 217 wide 12x9V datasets were accepted for analysis. Scleral ring disc margin points and corresponding disc area measurements were detected automatically using the commercially released software in DRI OCT Triton (Fastmap v10.05). Bruch’s membrane opening disc margin points and disc area measurements were automatically detected using an in-house developed software routine. The calculated disc areas based on the two methods were compared, and in addition, the repeatability standard deviation statistics were calculated using a nested ANOVA model.
For the scleral ring based method, the average disc area measurements and associated standard deviations were 1.55±0.045 mm2 (disc 6x6H), 1.68±0.041 mm2 (wide 12x9H), and 1.69±0.056 mm2 (wide 12x9V) for the normal group and 1.76±0.038 mm2 (disc 6x6H), 1.77±0.040 mm2 (wide 12x9H), and 1.86±0.062 mm2 (wide 12x9V) for the glaucoma group. For the Bruch’s membrane opening based method, the average disc area measurements and associated standard deviations were 1.77±0.057 mm2 (disc 6x6H), 1.87±0.060 mm2 (wide 12x9H), and 1.79±0.042 mm2 (wide 12x9V) for the normal group and 2.05±0.064 mm2 (disc 6x6H), 2.09±0.058 mm2 (wide 12x9H), and 2.06±0.076 mm2 (wide 12x9V) for the glaucoma group.
Both disc detection methods demonstrate good repeatability and show potential to distinguish glaucomatous eyes from normal eyes. Meanwhile, the Bruch’s membrane opening based disc area is larger than scleral ring based disc area as expected when border tissue is present.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.
This PDF is available to Subscribers Only