Standardized protocol for measuring ciliary body parameters offers reliable method for quantifying ciliary body morphology in ultrasound biomicroscopy images

Abstract

Purpose : Currently, there is no reliable method for analyzing the ciliary body (CB) in anterior segment (AS) images captured with ultrasound biomicroscopy (UBM). This study utilized a standardized protocol for measuring 6 CB parameters in transverse UBM images of the AS which capture an entire row of ciliary processes (CP). We hypothesized that analyzing the CB in transverse images focused at the CB will offer improved intra-observer reliability (IOR) and interobserver agreement (IOA) for analyzing CB parameters compared to previous literature.

Methods : Transverse UBM images of the CB were obtained from 26 eyes in 21 pediatric patients without AS pathology undergoing examination under anesthesia. 6 parameters in the CB were identified and measured using ImageJ software. Four trained medical student observers measured the 6 parameters in each image twice. IOR was determined by calculating the coefficient of variation (CV) and repeatability coefficient (r). IOA between each pair of observers and among all 4 observers was determined for each parameter using intraclass correlation coefficient (ICC), followed by Bland-Altman plots.

Results : All 6 CB parameters demonstrated high IOR, with ciliary body thickness (CBT), ciliary process thickness (CPT), and ciliary process length (CPL) measurements being the most consistent for each observer. The CV for each observer ranged from 1.43-14.8%. For each parameter, the range of calculated r values for the 4 observers is presented in Table 1. IOA was also high for all 6 parameters, with CPT, CPL, and ciliary process integrated density (CPID) measurements having the highest level of agreement among the 4 observers. ICC for CPT, CPL, and CPID was 0.96, 0.95, and 0.96, respectively. For the other parameters, ICC ranged from 0.84-0.91.

Conclusions : Utilizing transverse UBM images dedicated to capturing the CB allows for more consistent quantitative analysis of CB parameters. There is both improved IOR and IOA when compared to prior studies which analyzed CB parameters with axial or longitudinal UBM images. With a standardized protocol for measurement of CB parameters, quantifiable differences in CB morphology can be reliably measured in healthy individuals.

This is a 2020 ARVO Annual Meeting abstract.