December 2002
Volume 43, Issue 13
ARVO Annual Meeting Abstract  |   December 2002
Identifying Visual Field Progression with A Fuzzy Logic Model
Author Affiliations & Notes
  • DC Hoffman
    Jules Stein Eye Institute UCLA School of Medicine Los Angeles CA
  • DE Gaasterland
    Georgetown University Washington DC
  • J Caprioli
    Jules Stein Eye Institute UCLA School of Medicine Los Angeles CA
  • Footnotes
    Commercial Relationships   D.C. Hoffman, None; D.E. Gaasterland, None; J. Caprioli, None. Grant Identification: NIH EY12738
Investigative Ophthalmology & Visual Science December 2002, Vol.43, 2169. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      DC Hoffman, DE Gaasterland, J Caprioli; Identifying Visual Field Progression with A Fuzzy Logic Model . Invest. Ophthalmol. Vis. Sci. 2002;43(13):2169.

      Download citation file:

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

  • Supplements

Abstract: : Purpose: To develop a method of detecting visual field progression with fuzzy logic. Methods: Humphrey 24-2 and 30-2 visual field series from 293 eyes of 293 patients (mean follow up 8.3 ± SD 2.67 years, min 3.1, max 13.0) from 5 centers participating in the Advanced Glaucoma Intervention Study (AGIS) were analyzed. Series with less than 8 visual fields or series with an initial field with an AGIS visual defect score of 17 or more were excluded. When both eyes of a subject were eligible, one eye was randomly selected for enrollment. The series of visual fields for each eye were classified as progressing or stable according to the AGIS scoring system. Relative sensitivities of 5791 exams were averaged into 5 AGIS visual field clusters. Fuzzy logic is a rule-based method of artificial intelligence that attempts to balance the significance and precision of data. A fuzzy logic model for identifying visual field progression was created with subtractive clustering and 6 membership functions with the adaptive neuro-fuzzy inference system software (The MathWorks, Inc.). Two thirds of the data was trained and tested against the other third. This modeling was performed three times and the results were averaged and compared to results from a neural network. Results: The fuzzy logic model had a sensitivity of 75% and a specificity of 74% to detect visual field progression in a series compared with AGIS. A neural network had a sensitivity of 74% and a specificity of 85%. The agreement between the two was 72.6 % percent agreement, kappa =.4. Conclusion: Visual field progression can be detected by a novel approach called fuzzy logic and nearly as well as a neural network can. Further research will identify how to optimize fuzzy logic functions to obtain better results to identify visual field progression.

Keywords: 624 visual fields 

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.