Abstract
Purpose: :
To use anterior-segment optical coherence tomography (AS-OCT) measurements to develop a new classification system to discriminate patients at risk for angle closure from normal subjects.
Methods: :
Subjects over 50 years with no previous ophthalmic problems were recruited from a health clinic in Singapore. All subjects underwent AS-OCT imaging and gonioscopy in dark conditions by two independent masked examiners.Eyes were classified as having angle closure if the posterior trabecular meshwork could not be seen in >180o on gonioscopy. An experienced examiner manually determined the scleral spur location on AS-OCT horizontal scan images, and the Zhongshan Angle Assesment Program (Guangzhou, China) automatically provided anterior segment measurements such as: angle opening distance (AOD), trabecular-iris space area (TISA), anterior chamber depth (ACD), anterior chamber width (ACW), lens vault (LV), iris thickness (IT), and iris curvature (IC). The quantitative measurements of 1000 randomly selected subjects were used to construct receiver operating characteristic (ROC) curves. The best diagnostic parameters were selected to construct models to assess the risk for angle closure. This classification system was subsequently tested on an independent random selected sample of 465 eyes from the same health clinic in Singapore. Only the right eyes with clearly visible scleral spurs were included in the analysis.
Results: :
Angle closure on gonioscopy was observed on 215 (21,5%) eyes on the 1,000 subjects group, and on 101 (21,58%) eyes on the 465 independent sample group. Three sectorial classification systems were developed: 1) central - including a combination of ACD, ACW, LV, IT, and IC measurements; 2) temporal and 3) nasal - including nasal and temporal AOD and TISA measurements. If anterior chamber measurements were over the cutoff points determined on ROC curves with specificity locked at 99% or 95%, the status of the classification system were "outside normal limits (ONL)" or "Borderline", respectively. An "ONL" or "Borderline" result in any of the three sectorial classification system yielded the best discriminatory performance for angle closure, with a sensitivity of 91.6% and a specificity of 84.6%. Similar diagnostic performance was observed when the classification system was applied to the independent sample (sensitivity of 83.2% and specificity of 88.8%).
Conclusions: :
This new AS-OCT classification system using a combination of anterior chamber measurements demonstrated good discrimination between normal and angle closure eyes in these population from Singapore.
Keywords: imaging/image analysis: clinical • imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • anterior chamber