Variability and Reproducibility of Optic Disc Property Measurements within Bruch’s Membrane Opening and Neural Canal Opening

Abstract

Purpose: The Bruch’s Membrane Opening (BMO) or the Neural Canal Opening (NCO) is commonly used to define optic disc margins. A statistical study was performed to reveal any variability of optic disc property measurements made within BMO compared to within NCO, and to investigate whether results were reproducible.

Methods: Normal human subjects were scanned with a 67 kHz Cirrus 5000 HD-OCT based Optical Microangiography (OMAG) prototype system (Zeiss, Dublin, CA; 245x245 raster cube scan pattern over a 2.4x2.4mm² area centered at the optic nerve head region). BMO and NCO margins were delineated based on structural en face images. In the first analysis, 2 observers each calculated the BMO and NCO aspect ratio and area. The difference of each corresponding measurements within the same margin between observers were compared. In the second analysis, variability of BMO and NCO geometric center, aspect ratio, and area was found using a paired t-test. Each eye served as its own control. In the third analysis, variability of BMO and NCO blood flow index (BFI) was found using a paired t-test. Each eye served as its own control. Two trials were conducted, each by different observers. Measurements statistically different were defined by p<0.05.

Results: 10 eyes from 6 normal human subjects (67.0±12.6 years) were recruited. In the first analysis, the difference of corresponding properties between observers was found not significant (p=0.55 and 0.37 for BMO aspect ratio and area, respectively). Similar results were found for the NCO aspect ratio and area (p= 0.50 and 0.68, respectively). In the second analysis, no statistically significant variability between each corresponding BMO and NCO properties was detected (p=0.43, 0.46, and 0.92 for center, aspect ratio, and area difference, respectively). In the third analysis, variability of BMO and NCO BFI was found not significant in both trials (p=0.20 and 0.59 for trial 1 and 2, respectively).

Conclusions: The first test indicated that properties measured within disc margins defined either by the BMO or NCO is reproducible between observers. The second test displayed no variability in corresponding properties measured within disc margins defined either by BMO or NCO. The last test suggested BFI measurements are also not affected by changing the disc margin definition. It also indicated the reproducibility of the measurement between observers.