Reproducibility of qualitative assessment of drusen volume in eyes with age related macular degeneration.

Abstract

Although an optical coherence tomography (OCT)-derived central drusen volume ≥0.03 mm3 has been found to be a risk factor for progression to late age-related macular degeneration (AMD), this parameter is not currently available on most OCT devices or acquisition protocols. The purpose of this study was to evaluate the ability of human graders to qualitatively assess drusen volume by inspection of OCT B-scans. 100 subjects (200 eyes) from the Amish Eye Study diagnosed with early or intermediate AMD underwent OCT imaging with both Cirrus OCT and Spectralis OCT. Drusen volume was automatically computed from the Cirrus OCT volumes using the Cirrus Advanced RPE Analysis software. Spectralis volume scans were reviewed by two independent, masked graders who were asked to determine whether the central drusen volume was ≥0.03 mm3. Cohen's kappa coefficients were computed to assess the agreement. After excluding 11 eyes with poor image quality and 5 eyes used for training of the graders, the remaining 184 eyes were included in this analysis. The agreement between the graders and the automated evaluation of drusen volume by the Cirrus OCT was excellent with K = 0.88 for grader 1 and K = 0.82 for grader 2. The agreement between graders was also excellent with a K = 0.88. The presence of a high central drusen volume can be assessed reliably by qualitative inspection of OCT B-scans. This approach may be useful in the assessment of risk for progression to late AMD.