Palpebral fissure length measurement: accuracy of the FAS facial photographic analysis software and inaccuracy of the ruler

J Popul Ther Clin Pharmacol. 2015;22(1):e9-e26. Epub 2015 Dec 2.


Background: Accurate fetal alcohol spectrum disorder diagnoses require accurate facial measurement. The Fetal Alcohol Syndrome (FAS) Facial Photographic Analysis Software was developed to overcome measurement error known to occur with ruler measurement of the PFL. Recent publications have queried the Software's accuracy.

Objectives: 1) Demonstrate the Software's ability to accurately measure a PFL from a 2-dimensional digital facial photograph. 2) Demonstrate the frequency and magnitude of error when the PFL is measured directly by clinicians using a ruler.

Methods: Objective 1: PFLs of mannequins were measured using the Software and a sliding digital caliper, with the latter serving as the gold-standard accurate measure. Mannequins allowed the caliper prongs to be placed directly on the landmarks that define the PFL. Objective 2: PFLs of 1,027 patients evaluated at the University of Washington FAS Diagnostic & Prevention Network were measured with the Software and directly by one or two clinicians using a ruler.

Results: Objective 1: The Software derived PFLs that were identical to or within 0.2 mm of the caliper measures. Objective 2: There was tremendous inter-rater variability in PFLs measured by clinicians using a hand held ruler. Seventy-seven percent of patients had their PFLs measured incorrectly (greater than 1 mm error) by at least one of the two clinicians using a ruler.

Conclusion: The FAS Facial Photographic Analysis Software measures the PFL with the same accuracy as a sliding digital caliper, as it was programmed to do. Direct measurement of the PFL with a ruler is very prone to error.

Publication types

  • Comparative Study
  • Evaluation Study
  • Research Support, N.I.H., Extramural
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adolescent
  • Adult
  • Anthropometry / instrumentation
  • Anthropometry / methods*
  • Child
  • Child, Preschool
  • Equipment Design
  • Eye Abnormalities / pathology*
  • Female
  • Fetal Alcohol Spectrum Disorders / pathology*
  • Humans
  • Image Processing, Computer-Assisted / methods*
  • Male
  • Manikins
  • Middle Aged
  • Observer Variation
  • Photography / instrumentation
  • Photography / methods*
  • Predictive Value of Tests
  • Reproducibility of Results
  • Software Validation*
  • Washington
  • Young Adult