The quantitative measurement of facial photographs, termed photogrammetry, is an important procedure in multiple fields including ophthalmic, dental, bioanthropological, forensic, medical and other morphological sciences. When photographs are not life size, accurate measurements may be made by converting pixel lengths to millimeters, using a conversion factor, set by an object of known length, such as a line gauge, within the image. When an appropriate scale is not present, the conversion factor may be set by using an anatomical feature of known size. If the anatomical feature holds a narrow range of variation, a sample mean of the feature may, in some cases, suffice as a suitable substitute for an individual’s ground truth dimension. Due to their small variance, sample means of corneal diameter and palpebral fissure length have previously been used as scales, however, the vast published datasets have not been synthesized to triangulate on central tendencies. Herein we review methods and the full breadth of published data for adults and subadults (last five years) to calculate weighted pooled means and standard deviations for each anatomical structure. For the corneal diameter, these were 11.84 mm and 0.79 mm respectively (n = 296,887 eyes). The weighted pooled mean and standard for the most commonly measured 2D palpebral fissure dimension were 27.4 mm and 3.2 mm respectively (n = 5,871 eyes). A pilot trial of the pooled palpebral fissure length values for improved focus distance estimation in craniofacial superimposition using PerspectiveX is presented.
Healy et al. (Mon,) studied this question.