Abstract
Abstract: :
Purpose: To investigate RNFL detection rate using OCT with the existing and a new algorithm. Methods: Eyes with glaucoma undergoing peripapillary circular OCT (Zeiss-Humphrey Systems Inc., Dublin, CA) scanning were enrolled in a prospective clinical trial. Three consecutive circular scans were obtained for each eye. The inner and outer RNFL borders were traced using the original OCT software and a new software program. The new algorithm searches for peaks on each sampling line (analogous to the A-scan line on sonography) instead of applying conventional thresholding techniques. Detection error was defined as the presence of five consecutive sampling lines with disrupted RNFL borders. Results: Sixty-seven eyes of 67 patients (total 201 images) were enrolled (mean age 57.0 ± 12.6 (SD) yrs., range 23 to 84 yrs.; average mean deviation (MD) -5.0 ± 6.2 dB, range -29.8 to 1.0 dB). Failure to detect the RNFL occurred in 0.5% (1 image) with the new algorithm and 23.4% (47 images) with the existing algorithm (p<0.001, Fisher's exact test). Both algorithms showed significant correlations of RNFL measurements with MD (rs=0.61 (existing algorithm), 0.63 (new algorithm); p<0.001, Spearman’s rank correlation). Conclusion: Detection of RNFL borders remains problematic for existing OCT algorithms. Second generation algorithms will ameliorate this problem.
Keywords: 356 clinical (human) or epidemiologic studies: systems/equipment/techniques • 430 imaging/image analysis: clinical • 432 imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound)