Reproducibility of topographic measurements of the optic nerve head with the Heidelberg Retina Tomograph using separate and imported contour lines

Purpose: To investigate whether topographical measurements of the optic nerve head are more reproducible when the contour line is redrawn for each image as opposed to imported using the Heidelberg Retina Tomograph (HRT). Introduction: Optic nerve head topography can be measured in vivo using Confocal Scanning Laser Ophthalmoscopes such as the HRT. Image analysis requires that the area of the optic disc is defined by subjectively drawing a contour line around Elschnig's ring. The HRT software can export this contour line between images for a subject, or the contour line can be redrawn. Method: Ten eyes of ten normal volunteers (40–60 years) were analysed with the HRT. Seven 10° images of the optic disc were taken by one observer for each randomly selected eye. Focus and scan depth settings were kept constant. All available stereometric parameters were analysed using three methods: (1) with a separately drawn contour line for each individual image, (2) with a contour line imported from the first image into subsequent images of the series, (3) by repeatedly redrawing the contour line for a single image from each eye. All contour lines were drawn by one observer. Mean co-efficients of variation were calculated for each parameter for the three methods. Results: The mean co-efficient of variation for the parameters ranged from 1.7–23.4% (method 1), 0–20% (method 2) and 1.6–13.4% (method 3). Conclusion: These results show that redrawn contour lines increase variability. Topographical measurements of the optic nerve head are more reproducible if the contour line is imported across subsequent images rather than redrawn for each individual image.