Clinical Performance of Semi-Automated Spectral-Domain Optical Coherence Tomography Angiography

<b>Background/Objectives</b>: To evaluate the clinical performance of two optical coherence tomography angiography (OCTA) devices, including a semi-automated device, with respect to image quality and pathology detection, with fluorescein angiography (FA) and indocyanine green angiography (ICGA) serving as the reference standards. <b>Methods</b>: In this prospective cross-sectional study, normal eyes and those with various retinal and choroidal pathologies were enrolled and underwent OCTA scanning using semi-automated 3D OCT-1 Maestro2 and Cirrus™ HD-OCT 5000 devices, as well as FA/ICGA imaging. OCTA scans and FA/ICGA images were independently graded for image quality and the visibility of prespecified anatomic vascular features, along with the presence or absence of pathology on the OCTA scans and the FA/ICGA images (within regions corresponding to the OCTA scan areas). Positive percent agreement (PPA), defined as the proportion of eyes in which the OCTA demonstrated pathology when the corresponding FA/ICGA showed pathology, and negative percent agreement (NPA), defined as the proportion of eyes in which the OCTA showed no pathology when the FA/ICGA also showed no pathology, were calculated. <b>Results</b>: In total, 38 normal eyes and 86 pathologic eyes were enrolled in the study. The majority of images for both devices were considered clinically useful. The PPA and NPA were high for both devices, indicating a good ability to identify disease when present and to rule it out when not present. <b>Conclusions</b>: The findings of this study suggest that the semi-automated Maestro2 and Cirrus have comparably good clinical performance, particularly with regard to accuracy when identifying vascular pathologies.