December 2002
Volume 43, Issue 13
ARVO Annual Meeting Abstract  |   December 2002
Automated Objective Characterization of Visual Field Defects in 3D
Author Affiliations & Notes
  • W Fink
    Department of Ophthalmology Doheny Eye Institute Keck School of Medicine University of Southern California Los Angeles CA
  • R Castano
    Machine Learning Systems Group Jet Propulsion Laboratory/Caltech Pasadena CA
  • Footnotes
    Commercial Relationships   W. Fink, None; R. Castano, None.
Investigative Ophthalmology & Visual Science December 2002, Vol.43, 240. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      W Fink, R Castano; Automated Objective Characterization of Visual Field Defects in 3D . Invest. Ophthalmol. Vis. Sci. 2002;43(13):240.

      Download citation file:

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

  • Supplements

Abstract: : Purpose: To provide an automated, objective characterization of visual field defects in three dimensions, enabling the creation of a database and the monitoring of scotoma development over time. Methods: Visual field defects (resulting from AION, optic neuritis, ARMD, glaucoma, and others) gathered using the 3D computer-automated threshold Amsler grid test, developed by Fink and Sadun ( and, were evaluated with respect to area of visual field loss as a function of contrast level, overall volume of visual field loss as compared to a "normal" hill-of-vision, and slope distribution of the scotoma boundaries. Results: Even for more complex distributed visual field defects such as in cases of optic neuritis, the area and volume of visual field loss can be automatically calculated, allowing monitoring the progression over time. Furthermore, the distribution of boundary slopes (slope defined by the percent of retinal contrast sensitivity loss over degrees of visual field) within a particular visual field defect gives a measure of the character of the scotoma: a sharply peaked distribution around a steep slope implies an absolute type of scotoma, whereas a sharply peaked slope distribution around a shallow slope implies a relative type of scotoma. The presence of steep and shallow slopes within a visual field defect leads to a multi-center distribution of slopes. Conclusion: The 3D computer-automated threshold Amsler grid test used for visual field testing provides several advantages over conventional perimetry including additional information through 3D rather than 2D depiction of scotomas and superior angular resolution (1 degree compared to typically 6 degrees). This allows for an unprecedented characterization of the 3D structure of scotomas, thus laying the foundation for both, creating a database and monitoring the development of scotomas typical for various diseases over time. Eventually this approach would enable the creation of an autonomous computer-based diagnosis system for visual field defects, that could assist physicians with an independent second opinion or provide expertise where otherwise not readily available.

Keywords: 624 visual fields • 368 contrast sensitivity • 486 neuro-ophthalmology: diagnosis 

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.