Purchase this article with an account.
Helen Koenigsman, Ruojin Ren, Hongli Yang, Stuart Gardiner, Juan Reynaud, Robert Kinast, Steven Mansberger, Brad Fortune, Shaban Demirel, Claude Burgoyne; Variability in Optic Disc Margin and Rim Assessment by Glaucoma Specialists using Color Stereophotographs. Invest. Ophthalmol. Vis. Sci. 2013;54(15):64.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To characterize inter-clinician (IC) variability in assigning disc margin and rim margin within 184 color stereophotos of glaucoma and glaucoma suspect eyes.
Photos of 184 eyes of 184 subjects were independently evaluated by 5 trained glaucoma specialists. Stereophoto pairs were viewed using a handheld stereoscope while the disc margin and rim margin were marked using a separate monitor and customized software. For each photo, the centroid of each clinician’s disc margin was calculated, and an average DMcentroid (n=5 clinicians) was determined. Photos were then colocalized to confocal scanning laser ophthalmoscopy (cSLO) infrared fundus reflectance images of the same eye so that the axis between DMcentroid and the fovea could be used to establish anatomic 30-degree sectors in each eye. (Fig 1) Radial distance measurements from DMcentroid to each clinician’s disc margin (disc radius) and rim margin (cup radius) were noted and used to calculate the rim width, disc area, cup area and rim area in each sector. IC variability for each parameter (disc radius, cup radius, rim width, disc area, cup area and rim area) within the 184 photos was assessed as the standard deviation of differences from the intra-eye inter-clinician mean within each sector.
IC variability for each parameter was substantial and regionally variable in these 184 eyes. Among the distance parameters, cup radius was more variable than disc radius in all 12 sectors with each being greatest nasally (see Fig 2 for polar plots of the SD of residuals for all parameters). Cup area was more variable than disc area superotemporally but less variable inferonasally due to the reduced extent of the cup in the inferonasal region. Variability in rim width comprises the effect of variability on both rim margin and disc margin, so is greater than either. Similarly rim area is more variable than either cup area or disc area since it incorporates both sources of variability.
IC variability in rim margin and disc margin assessment was substantial among the glaucoma specialists of this study leading to marked IC variability in rim width and rim area assessment. Reducing variance due to a well known source of anatomical variation is beneficial. A regionalization strategy based on the angle between the disc center and fovea is a useful approach for standardizing optic disc assessment by sectors.
This PDF is available to Subscribers Only