Investigative Ophthalmology & Visual Science Cover Image for Volume 61, Issue 7
June 2020
Volume 61, Issue 7
Free
ARVO Annual Meeting Abstract  |   June 2020
Estimating Visual Sensitivity from Single-Intensity Stimuli in Visual Field Tests for Patients with Glaucoma: Theory
Author Affiliations & Notes
  • Benjamin T Backus
    Vivid Vision, Inc., San Francisco, California, United States
  • James Blaha
    Vivid Vision, Inc., San Francisco, California, United States
  • Manish Gupta
    Vivid Vision, Inc., San Francisco, California, United States
  • Footnotes
    Commercial Relationships   Benjamin Backus, Vivid Vision, Inc. (E), Vivid Vision, Inc. (I), Vivid Vision, Inc. (P); James Blaha, Vivid Vision, Inc. (E), Vivid Vision, Inc. (I), Vivid Vision, Inc. (P); Manish Gupta, Vivid Vision, Inc. (E), Vivid Vision, Inc. (I), Vivid Vision, Inc. (P)
  • Footnotes
    Support  None
Investigative Ophthalmology & Visual Science June 2020, Vol.61, 3895. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Benjamin T Backus, James Blaha, Manish Gupta; Estimating Visual Sensitivity from Single-Intensity Stimuli in Visual Field Tests for Patients with Glaucoma: Theory. Invest. Ophthalmol. Vis. Sci. 2020;61(7):3895.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : In visual field testing a distinction is often made between “threshold” tests, in which stimulus intensities vary according to a staircase procedure, and “suprathreshold” tests, in which stimulus intensity is fixed. But varied intensities are not needed to estimate thresholds, because psychometric functions are graded. A stimulus of fixed intensity will be seen 0-100% of the time depending on threshold. Thus, it may be possible to monitor the progression of glaucoma using a fixed-intensity test. Internal (neural) noise is lower when detecting predictable targets, and patients might prefer fixed-intensity tests. How well does a fixed-intensity scheme work to monitor progression in glaucoma?

Methods : Matlab was used to simulate patient performance for fixed- and variable-intensity test algorithms for a 24-2 layout (54 locations). Model parameters were fixed-stimulus intensity, number of trials, and slope of the psychometric function. Simulated patients had various patterns of glaucomatous vision loss. Threshold was estimated at each location by maximizing likelihood. Mean sensitivity was the mean threshold across the 54 locations, and power to detect progression was computed by Monte Carlo simulation.

Results : For patients with diffuse mild loss, the variable-intensity algorithm performed better because the fixed-intensity test had too few misses for thresholds to be estimated. However, in patients with significant loss, both mean sensitivity estimation and progression detection were almost as good for the fixed-intensity algorithm as for the variable-intensity algorithm.

Conclusions : For patients with manifest glaucoma, a fixed-intensity test is almost as efficient as a variable-intensity test when trying to estimate mean sensitivity or detect progression. Given that real patients may prefer fixed-intensity tests, and that predictable stimuli may promote consistent performance, fixed-intensity tests could be optimal for monitoring visual fields in patients with glaucoma.

This is a 2020 ARVO Annual Meeting abstract.

 

Est. thresholds, for 4 true thresholds (green arrows). Model psychometric function slope sigma=10 dB (cum norm). Fitted slope 5=dB, fixed intensity=15 dB, varied intensity staircase start=15 dB, step-size=7.5 dB.

Est. thresholds, for 4 true thresholds (green arrows). Model psychometric function slope sigma=10 dB (cum norm). Fitted slope 5=dB, fixed intensity=15 dB, varied intensity staircase start=15 dB, step-size=7.5 dB.

 

Top: Est. threshold as a function of true threshold. Bottom: Bias (left) and variability (right) as a function of no. trials per location (x-axis). Panels are patients with different patterns of loss.

Top: Est. threshold as a function of true threshold. Bottom: Bias (left) and variability (right) as a function of no. trials per location (x-axis). Panels are patients with different patterns of loss.

×
×

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.

×