Abstract
Purpose :
Measurement of the iridocorneal angle with ultrasound biomicroscopy (UBM) or optical coherence tomography (OCT) can provide objective assessment of the risk or progression of angle closure. The most widely used parameter is the Angle Opening Distance (AOD) which measures the distance from the corneal endothelium to anterior iris. AOD500 is the measurement at 500 microns from the sclera spur. The measurement can be performed manually or with a semi-automated (SA) tool on the Absolu ultrasound system (Quantel Medical). The purpose of this study is to compare measurements made with manual versus SA methods as well as comparing inter-observer measurements.
Methods :
We conducted a retrospective review of UBM exams of 12 eyes of 6 patients referred for evaluation of anatomically narrow angles. Cineloops of up to 100 images were obtained using the Lin50 probe. Two experienced ultrasonographers independently chose three images optimally displaying the angle and scleral spur in four quadrants (nasal, superior, temporal and inferior) of each eye to obtain AOD500 measurements both manually and separately with SA software. Measurement time was recorded with both methods. Measurement variation within and between quadrants by each technique was performed with Bland-Altman analysis and interobserver variation determined.
Results :
A total of 138 images were evaluated; mean AOD500 was 0.275 mm (manual) and 0.277 mm (SA). Intra-quadrant standard deviation (SD) was 0.037mm (manual) and 0.044 mm (SA). Inter-quadrant SDs were 0.076 and 0.079 respectively. Manual versus SA correlation coefficient was 0.93. Inter-observer correlation coefficients were 0.72 mm (manual) and 0.59 mm (SA). The inter-observer root mean square difference was 0.077 mm (manual) and 0.100 mm (SA). Coefficients of variation were 0.30 (manual) and 0.41 (SA).
Conclusions :
There was little variation between manual and semi-automated measurement techniques. SDs were slightly greater for measurements obtained by the SA method but not statistically significant. The chief source of uncertainty is identification of the scleral spur, which must be performed manually in either method. SA determinations were more time-efficient utilizing 57% less time than manual; including identification of scleral spur the overall time saving was 23%.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.