Development and validation of a computerized expert system for evaluation of automated visual fields from the Ischemic Optic Neuropathy Decompression Trial
2006

Computerized System for Analyzing Visual Fields in NAION Patients

Sample size: 284 publication 10 minutes Evidence: moderate

Author Information

Author(s): Steven E. Feldon, Lori Levin, Roberta W. Scherer, Anthony Arnold, Sophia M. Chung, Lenworth N. Johnson, Gregory Kosmorsky, Steven A. Newman, Joanne Katz, Patricia Langenberg, P. David Wilson, Shalom E. Kelman, Kay Dickersin

Primary Institution: University of Rochester School of Medicine

Hypothesis

Can a computerized system accurately classify visual field defects in patients with non-arteritic anterior ischemic optic neuropathy (NAION)?

Conclusion

The IONDT developed a rule-based computerized system that consistently defines pattern and severity of visual fields of NAION patients for use in a research setting.

Supporting Evidence

  • The computerized system showed good agreement with expert classifications for 91 of 95 visual fields.
  • 63.6% of patients with progressive NAION showed worsening visual fields within 30 days.
  • Expert panel agreement on classifications was only 69% for the validation set.

Takeaway

Researchers created a computer program to help doctors understand vision problems in patients with a specific eye condition, and it worked pretty well.

Methodology

The study used visual fields from 189 non-IONDT eyes with NAION to develop a computerized classification system, which was validated against expert panel classifications.

Potential Biases

The study's reliance on expert panel definitions may introduce bias, as experts may have differing opinions on visual field classifications.

Limitations

The expert panel did not reach complete agreement on a sizable proportion of defect classifications, indicating variability in human classification.

Participant Demographics

Patients aged 50 years or older with symptoms of NAION.

Statistical Information

P-Value

0.0003

Confidence Interval

95% CI

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.1186/1471-2415-6-34

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication