Facial aging measurement bench

One face, four questions any facial aging measure has to answer. Cohort references are recomputed from the study data for the exact clock each row runs, on the same 22,866 same-day photograph groups used in the paper.

Date of birth optional

Age acceleration is defined against chronological age, which a photograph does not carry. Enter a date and the residual is computed both ways, side by side.

Age used, sex –

Cohort percentile of face-attributable risk –

A rank within the study cohort, not an absolute risk. The paper supports a population statistic, not an individual prognosis.

1 · Every clock, one face

2 · Measurement noise

–

SD across 0 recent frames of the same, unchanged face

3 · The acceleration trap

4 · Photographic era

as shot today –

same face, 1975 film –

Ages outside the cohort range are refused rather than extrapolated. No image leaves this machine and nothing is stored. The mortality model is not validated for individual prognostic use.