The Effect of Scleral Spur Identification Methods on Structural Measurements by Anterior Segment Optical Coherence Tomography

Abstract

To assess methods for and variations in identifying the scleral spur (SS) position in anterior segment optical coherence tomography. In images of 51 eyes (patients) with open and closed anterior chamber angles, we compared the success rate and the variability of 3 approaches for identifying the SS: the ciliary muscle (CM), bump, and Schwalbe line (SL) methods using mixed effects regression models. The effect of incremental variation in SS position on anterior chamber parameters using the Anterior Segment Analysis Program (ASAP) was analyzed in 8 images. Automated ASAP measurements were compared with manual ImageJ measurements in 46 images. The SS could be identified in 98% of images by each observer using the 3 methods in combination. The SL and CM approaches more successfully identified the SS (82% and 81% success, respectively) than the bump method (59%, P<0.001). The intraobserver, interobserver, and intermethod variabilities of the CM and bump methods were superior to those of the SL method. The SS was more likely to be identified in open angle than angle closure eyes (OR=2.26, P=0.03) and brown eyes were less likely than blue eyes (OR=0.36, P=0.04). Movement of SS position resulted in substantial differences in the angle parameters and iris concavity ratio, whereas iris area and volume were less affected. The CM method was the most successful and least variable method of SS marking, which was more difficult in narrow angle and brown eyes. Variability of SS placement had a large effect on angle parameters and iris concavity ratio.