Anatomical Placement of the Human Eyeball in the Orbit—Validation Using CT Scans of Living Adults and Prediction for Facial Approximation*

Abstract

Accuracy of forensic facial approximation and superimposition techniques relies on the knowledge of anatomical correlations between soft and hard tissues. Recent studies by Stephan and collaborators (6,8,10) reviewed traditional guidelines leading to a wrong placement of the eyeball in the orbit. As those statements are based on a small cadaver sample, we propose a validation of these findings on a large database (n = 375) of living people. Computed tomography scans of known age and sex subjects were used to collect landmarks on three-dimensional surfaces and DICOM with TIVMI. Results confirmed a more superior and lateral position of the eyeball relatively to the orbital rims. Orbital height and breadth were used to compute regression formulae and proportional placement using percentages to find the most probable position of the eyeball in the orbit. A size-related sexual dimorphism was present but did not impact on the prediction accuracy.

The Validity of Ear Prediction Guidelines Used in Facial Approximation*,†,‡

“…Tests of previously published face prediction rules have been conducted for many of the major facial regions (eyes, nose, and mouth), except for the ear. This study fills that gap and its results are consistent with a general pattern observed in other studies that traditionally recommended face prediction rules are not well supported by the scientific data (10–18,61). …”

“…The inaccuracies of previously published soft tissue prediction guidelines are not ambiguous or limited to few specific facial regions. Similar findings have been found on numerous occasions and often by independent teams of investigators, see, for example, results on eyeball position (13–15,61,62). In this context, the poor performance of overarching face prediction methods (see, e.g., [26,63–65]) is not surprising and mandates improvements to facilitate accurate face prediction.…”

A geometric morphometric evaluation of the Belanglo ‘Angel’ facial approximation

“…With regards to eye spacing, the facial approximation also slightly underestimated eyeball width [63], and did not fully comply with research findings that the eyeball is displaced from the orbital centre by, on average, 1.4 mm superiorly and 2.3 mm laterally [64,65]. This was in part due to the cited studies involving a small number of cadaveric subjects of advanced age, but a large validation study involving 375 living adults [63] has since verified the findings. The decision to only displace the eyeballs by 1 mm superiorly and laterally, however, was in the main part because a greater displacement did not agree with the verified methods used to estimate mouth width.…”

Facial soft tissue thicknesses: Noise, signal, and P

“…P's overemphasis in the FSTT literature partly arises from the overuse of univariate statistics for multivariate data [3,89], where P-value results in studies tend to be extensively listed / tabulated. This bypasses a more holistic analysis gained by multivariate analysis and strength of association measures, such as eta-square with MANOVA ( 2 = SS Effect / SS Total ; where SS = sum of squares).…”