Repeatability of surface height, signal width, and edema index measurements of the HRT II in normal and diabetic subjects.

Abstract

Abstract: : Purpose: The Heidelberg retina tomograph II is a confocal laser scanning ophthalmoscope used for three dimensional imaging of the macula. Unlike the experimental version of the ophthalmoscope (HRT), the reliability of the instrument in assessing the macula has not been demonstrated. The purpose of the current study was to determine the repeatability of its measurements in normal and diabetic subjects. Methods: Three scans of each of 10 normal eyes and a minimum of 2 scans of each of 10 diabetic eyes were obtained within a single day. In addition, a second set of each of the 10 normal eyes was obtained on a different day. As a measure of repeatability, the coefficient of variation of the mean surface height, signal width, and edema index measurements bound by a 0.25 mm, 0.50 mm, and 0.75 mm radius circle were calculated. Moreover, the 95% confidence intervals of the difference between single as well as the average of three measurements for normal subjects on two different days were calculated from the pooled standard deviations. Results: The differences in variability of the measurements bound by the various circle sizes are statistically insignificant for both normal and diabetic subjects. Nonetheless, the values of the coefficient of variation for normal subjects and a 0.50 mm radius circle are 17%, 6%, and 6% for the mean surface height, signal width, and edema index, respectively, while the same values are 14%, 11%, and 11% for diabetic subjects. The 95% confidence interval of the difference between single observations in normal subjects is ± 106 µm, ± 80 µm, and ± 0.10 for mean surface height, signal width, and edema index, respectively. When three scans are combined, the corresponding values are ± 76 µm, ± 57 µm, and ± 0.07. Conclusions: The HRT II demonstrates good repeatability both for normal and diabetic subjects and its measurements are independent from the size of the examined regions. The repeatability of the mean signal width and edema index is slightly better for normal than diabetic subjects, and the combination of three scans provides a more reliable assessment of the macula in normal subjects.

Keywords: imaging/image analysis: clinical • macula/fovea • retina