Skip to main content

  • Prevalence, Types, and Severity of OCTA Artifacts

    By Lynda Seminara
    Selected and Reviewed By: Neil M. Bressler, MD, and Deputy Editors
    Add to My Bookmarks

    Journal Highlights

    JAMA Ophthalmology, February 2020

    Download PDF

    Artifacts on optical coherence tomo­graphic angiography (OCTA) images can lead to misinterpretation of the scans in clinical trials and real-world settings. Holmen et al. evaluated the prevalence and types of these artifacts and assessed their impact on image reliability. They found that more than half the study images had an artifact affecting the reliability of quantitative output. The most common artifacts were shadow, defocus, and movement.

    For this cross-sectional study, the authors acquired OCTA images from multiple multicenter clinical trials and submitted them to the Wisconsin Fun­dus Photograph Reading Center during a three-year period (2016-2018). The images had been captured with com­mercially available OCTA systems in accordance with the appropriate scan protocol (3 × 3 mm or 6 × 6 mm). The specific types of artifacts were assigned a severity grade, based on involvement of cross-sectional OCT and the area of OCT grid affected. Primary outcomes were the prevalence and severity of arti­facts, as well as area under the receiver operating characteristic curve (AUC) of quality indices with image reliability.

    Overall, 406 OCTA images (from 234 eyes) were included in this study. Of these, 221 (54.4%) were 6-mm scans, and 185 (45.6%) were 3-mm scans. At least one artifact was detect­ed in 395 images (97.3%). An artifact severe enough to affect the reliability of quantitative output was present in 217 images (53.4%). The most common artifacts were shadow (26.9%), defocus (20.9%), and movement (16%). Al­though artifact prevalence did not vary by imaging system or scan protocol, the types of artifacts differed. The com­mercially recommended quality index thresholds had an AUC of 0.80 to 0.83, sensitivity of 97% to 99%, and specific­ity of 37% to 41% for reliable images.

    The findings suggest that artifacts affecting quantitative output of OCTA devices are highly prevalent at this time. (Also see related commentary by David Sarraf, MD, and SriniVas Sadda, MD, in the same issue.)

    The original article can be found here.