QUALITY ISSUES IN INTERPRETATION OF OPTICAL COHERENCE TOMOGRAMS IN MACULAR DISEASES

Abstract

To analyze the scan characteristics associated with poor-quality Stratus optical coherence tomograms submitted to a reading center for multicenter clinical trials. Data from evaluation of 6,741 fast macular thickness map reports from trials involving age-related macular degeneration (AMD), diabetic macular edema, and retinal vein occlusion were analyzed. Optical coherence tomograms with an erroneous centerpoint thickness needing manual remeasurement (MR) were categorized as being of poor quality. The frequency of MR and the artifacts associated were analyzed by disease type, underlying retinal morphology, and severity of retinal thickening. MR was performed in 2,027 (30%) optical coherence tomograms. AMD had the highest frequency of MR (54.9%), followed by retinal vein occlusion (23.9%) and diabetic macular edema (16.3%). Boundary line errors were the most common artifact across all disease types (61.3% of scans requiring MR) and increased with increasing retinal thickness. Decentration artifact was seen in 15.4% of scans requiring MR. The median absolute difference between machine and manually measured centerpoint thickness assessed in a subset of 84 scans was 75.5 microm. Artifacts causing erroneous reported centerpoint thickness are common. Identifying clues that indicate suboptimal quality of optical coherence tomography (OCT) images are important to avoid erroneous interpretation of OCT data in clinical trials.