Software-based evaluation of human age: a pilot study.

PURPOSE: This pilot study was to assess a smartphone application regarding its use as an objective evaluation tool for subject age in comparison to human raters and to identify potential factors influencing the estimation of age.
MATERIALS AND METHODS: Ten Caucasian participants (six females, four males, mean age 42.1 ± 22.6 years) were randomly chosen, and frontal facial pictures of each participant were taken. The smartphone application PhotoAge (Version 1.5, ©2012, Percipo Inc., San Francisco, CA, USA) was used to evaluate the age of the participants. For comparison, 100 randomly selected raters (60 females, 40 males, mean age 29.3 ± 1.3 years) were asked to evaluate the age of the same participants. The influence of participants' facial expression, age, and sex as well as raters' age, sex, and profession was investigated as well. Statistical analyses (linear mixed models with random intercepts; least square means, confidence interval 95%; p < 0.05) were implemented.
RESULTS: PhotoAge resulted in a mean age of 43.1 ± 18.2 years, with a difference from the true mean age of 1.0 ± 8.2 years (p = 0.5996). The evaluation by the raters revealed a mean age of 41.5 ± 19.0 years, with a difference from the true mean age of -0.6 ± 8.5 years (p = 0.6078). There was no statistical significance between the two groups (p = 0.2783).
CONCLUSION: The evaluation of age with the software application PhotoAge seems to be a reliable procedure with comparable results to human raters. CLINICAL SIGNIFICANCE: This study gives a better understanding about the reliability of a software-based evaluation tool for age and identifies factors (e.g., the visibility of the teeth) potentially affecting the estimation of age. Naturally looking teeth seem to have no influence on the evaluation of a person's age. Thus, the application of this specific application for dental purposes is questionable; however, in forensics, it might be a valuable tool for estimating a person's age.