Abstract
Abstract: :
Purpose: To determine the frequency and type of OCT scan artifacts, and whether these artifacts depend on patient demographics and underlying retinal pathology. Methods:OCT was performed using the fast macular thickness map mode on the OCT Stratus. The OCT topographic thickness maps and the six radial scans from which the maps were derived were examined. The presence and type of scan artifacts, the reason for the artifacts, and clinical data were entered into an electronic database. Frequencies and percentages for all variables were computed. Logistic regression was used to relate the presence of OCT scan artifacts simultaneously with patient demographic data and underlying diagnosis. Results: Scans from 171 eyes were analyzed. Retinal artifacts, though not observed in normal eyes, were identified frequently in eyes with macular pathology (p<0.05). Artifacts were observed in 43.8% of all scans, and of these, the retinal thickness measurement was adversely affected in 62.2%. Six types of artifacts were observed: computer software misidentification of the inner and outer retina (29.8% and 24%, respectively), degraded image (11.7%), scan off–center (5.3%), scan "out–of–register" (3.5%), and "cut–off" scan edges (2.3%). Artifact affecting the surface map center thickness measurement occurred more frequently in eyes with one or more macular diagnoses than in those without such diagnoses (p=0.023). Eyes with neovascular age–related macular degeneration, macular holes, or non–proliferative diabetic retinopathy contained certain types of artifact more frequently than those eyes without these specific diagnoses (p<0.05) or those eyes without any diagnoses (i.e. normal eyes) (p<0.05). Misidentification of the inner and outer retina occurred more frequently in eyes with neovascular age–related macular degeneration than in those without (p=0.023 and p<0.001, respectively), as did "off center" artifacts (p=0.03). Misidentification of the inner retina occurred more frequently in eyes with macular hole than in eyes without this condition (p=0.043). Misidentification of both the inner and outer retina occurred more frequently in eyes with posterior vitreous detachment than in normal eyes (p=0.028 and p=0.019, respectively). Conclusions: OCT scan artifacts are seen frequently, affect retinal thickness measurements in a high proportion of cases, and are diagnosis–dependent. In many cases, artifacts are preventable. Recognition of these artifacts will improve retinal thickness measurement accuracy and will prevent faulty treatment decisions that are based on inaccurate retinal thickness measurements.
Keywords: imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • imaging/image analysis: clinical • retina