July 2019
Volume 60, Issue 9
Free
ARVO Annual Meeting Abstract  |   July 2019
Constructing an OCT parameter to maximize correspondence with visual field mean deviation
Author Affiliations & Notes
  • Ou Tan
    Ophthalmology, Oregon Health & Science Univ, Portland, Oregon, United States
  • Liang Liu
    Ophthalmology, Oregon Health & Science Univ, Portland, Oregon, United States
  • Xinbo Zhang
    Ophthalmology, Oregon Health & Science Univ, Portland, Oregon, United States
  • David Huang
    Ophthalmology, Oregon Health & Science Univ, Portland, Oregon, United States
  • Footnotes
    Commercial Relationships   Ou Tan, Optovue (P); Liang Liu, None; Xinbo Zhang, None; David Huang, Optovue (F), Optovue (I), Optovue (P), Optovue (R)
  • Footnotes
    Support  NIH grants R01EY023285, R21EY027007, P30 EY010572, Unrestricted grant from Research to Prevent Blindness
Investigative Ophthalmology & Visual Science July 2019, Vol.60, 3923. doi:https://doi.org/
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Ou Tan, Liang Liu, Xinbo Zhang, David Huang; Constructing an OCT parameter to maximize correspondence with visual field mean deviation. Invest. Ophthalmol. Vis. Sci. 2019;60(9):3923. doi: https://doi.org/.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose : To construct an OCT parameter that has maximal linear correlation and agreement with visual field (VF) mean deviation.

Methods : Participants in the Advanced Imaging for Glaucoma (AIG) study was screened for those with 4 consecutive visits with complete OCT and VF data of acceptable quality. Optic nerve head (ONH) OCT scans were obtained with a spectral-domain OCT system (RTVue). The retinal nerve fiber layer (NFL) thickness profile at the 3.4-mm diameter was evenly divided into 16 sectors. The NFL sector averages were transformed to a logarithmic decibel (dB) scale, then averaged using VF area weighting. The NFL weighted logarithmic average and VF-MD were averaged over 4 visits and then a quadratic fit was performed. The fit was used to transform the NFL measurement into the NFL-MD parameter. The modified Hodapp-Parrish-Anderson criteria was used to classify glaucoma severity.

Results : Two hundreds and forty five normal eyes, 411 pre-perimetric glaucoma (PPG) eyes, and 283 perimetric glaucoma (PG) had suitable data for analysis. The NFL-MD had higher correlation (Pearson R = 0.67) with VF-MD than overall NFL thickness (Pearson R = 0.52). The difference statistically significant (p<0.001). NFL-MD had significantly higher sensitivity than overall NFL thickness in detecting PPG (0.17 vs 0.10, p <0.001) and PG (0.62 vs 0.43, p = 0.001) at the 99% specificity level. NFL-MD and overall NFL thickness had similar within-visit repeatability (0.988 vs 0.988) and between visit reproducibility (0.978 vs 0.968). Bland-Altman analysis (Figure 1) showed good agreement between NFL-MD and VF-MD for PPG and mild PG. But in moderate to advanced PG NFL-MD tend to underestimate glaucoma severity in comparison to VF-MD.

Conclusions : We constructed a glaucoma parameter called NFL-MD that is based on OCT NFL profile but has the same dB scale. Compared to the overall NFL thickness, the NFL-MD has improved diagnostic sensitivity and correlation with VF-MD. NFL-MD may improve the management of glaucoma by giving clinicians an objective metric to evaluate glaucoma severity and progression that could be interpreted as an estimate of the highly familiar and validated VF-MD. However, agreement with VF-MD is limited in moderate and advanced glaucoma. Validation of this new metric in an independent dataset is needed.

This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.

 

Figure 1.Agreement between NFL-MD and VF-MD; SD: standard deviation of the difference, CI: confidence interval, unit: dB

Figure 1.Agreement between NFL-MD and VF-MD; SD: standard deviation of the difference, CI: confidence interval, unit: dB

×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×