Test‐retest reliability on measuring anterior chamber parameters with a novel automatic algorithm using anterior segment optical coherence tomography

Abstract

AbstractPurpose To describe the test‐retest variability in anterior chamber and iris parameters obtained on anterior segment optical coherence tomography (AS‐OCT) when measured with a novel automatic Location‐Based Graph Search segmentation algorithm.Methods Patients with open angles and primary angle closure on gonioscopic finding were invited for anterior chamber (AC) angle imaging with the Visante AS‐OCT. 3 scans of each eye were acquired in darkness with the patients repositioning their heads after each scan. The images were analysed with a custom software which used the iris rim as the landmark instead of the sclera spur. The parameters analyzed included angle opening distance and angle recess area at 750μm from the angle (AOD750 & ARA750), AC depth and width, lens vault, iris curvature, thickness, length and angle recess area. The test‐retest intrasession variability, coefficient of variation (COV) and intraclass correlation coefficient (ICC) were calculated to evaluate repeatability and reproducibility.Results 17 eyes with closed angles and 15 eyes with open angles were recruited. The ICC was >0.75 for most parameters of eyes with open angles and COV was <15%. For narrow angled eyes, the ICC was >0.75 for AC depth and width, lens vault, iris length and area, left AOD750 and ARA750 and angle recess, and <0.75 for other parameters. The COV for AC depth and width, iris length and area was <10% and >10% for other parameters.Conclusion There is test‐retest variability in AC parameters measured with the Visante AS‐OCT. The variability is greater in eyes with narrow angles but reliable in eyes with open angles. Caution must be exercised when using the AS‐OCT to follow up patients with narrow angles.