Abstract
Purpose:
Glaucomatous change, the most common form of optic nerve head (ONH) disease and main cause for vision impairment in aging populations, is intricately linked to disc size. Yet, current methods assessing the ONH are highly subjective and may be unreliable. The disc damage likelihood scale (DDLS), current “gold standard” in glaucoma assessment, relies on size evaluation most commonly obtained by slit lamp exam. We aimed to develop improved methods for this initial step that may aid accurate classification of the ONH and improve glaucoma diagnosis.
Methods:
The optic nerve head was visualized by four standard methods: (1) Haag-Streit slit lamp BQ900 in combination with a Volk Super 66D lens (SL), (2) Kowa nonmyd™ WX3D retinal camera at 45° in normal mode (RP), (3) Heidelberg Retina Tomograph (HRT), and (4) Cirrus HD-OCT. Automated calculation with the Cirrus HD-OCT ONH and RNFL OU Analysis function was compared to subjective measurements obtained by experienced clinicians. Statistical analysis was performed to calculate reproducibility and accuracy of the methods.
Results:
Intraclinician variation using different methods was typically around 10%, with a maximum of 56%. Interclinician measurements obtained from Cirrus HD-OCT images were most reproducible (variance < 5%) and highly correlated with automated values (R2=0.97, p<0.01), while HRT based disc size showed poor correlation with the automated system (R2=0.80, p<0.01). RP and SL both correlated equally well with the Cirrus ONH size (R2=0.90, p<0.01), but disc size was, on average, was smaller with RP (0.15±0.14mm2) and larger with SL (0.38±0.17mm2) resulting in a potential overestimated or underestimation of damage, respectively. While interclinician variance was comparable between HRT and RP (≤28%), SL assessments appeared not easily reproducible (≤58% variance).
Conclusions:
ONH assessment based on imaging from the Cirrus HD-OCT resulted in the least stratification between operators and the best correlation to the automated calculation, constituting the most robust method. However, the required equipment is highly specialized and costly. Based on current results, we propose use of retinal photography as a sustainable alternative. Thus, our results provide a basis to adjust current methods and adopt a unified system for ONH assessments to improve diagnostics for glaucoma and other ONH diseases.
Keywords: 627 optic disc •
465 clinical (human) or epidemiologic studies: systems/equipment/techniques •
550 imaging/image analysis: clinical