Reliability of Periocular Anthropometry: A Comparison of Direct, 2-Dimensional, and 3-Dimensional Techniques

BACKGROUND: Three-dimensional (3D) imaging has become increasingly popular in aesthetic surgery. However, few studies have emphasized its application in the periocular region. OBJECTIVE: To provide evidence supporting the reliability of generalizing periocular measurements obtained using caliper-derived direct anthropometry and 2-dimensional (2D) photogrammetry to 3D stereophotogrammetry. MATERIALS AND METHODS: Periocular surfaces were captured using a stereophotogrammetry system for 46 normal Caucasian individuals. Twenty-two periocular variables were directly, 2-dimensionally, and 3-dimensionally measured. Reliability of these measurements was evaluated and compared with each other. RESULTS: The results revealed that, for direct (intra-rater reliability only), 2D, and 3D anthropometry, overall intra-rater and inter-rater intraclass correlation coefficient estimates were 0.88, 0.99 and 0.97, and 0.98 and 0.92, respectively; mean absolute differences were 0.84 mm, 0.26 and 0.36 units, and 0.35 and 0.67 units, respectively; technical error of measurement (TEM) estimates were 0.85 mm, 0.25 and 0.36 units, and 0.32 and 0.65 units, respectively; relative error measurement estimates were 6.46%, 1.69% and 2.74%, and 1.67% and 5.11%, respectively; and relative TEM estimates were 6.25%, 1.62% and 2.78%, and 2.12% and 5.12%, respectively. CONCLUSION: Stereophotogrammetry and the authors' landmark location protocol yield very good reliability for a series of 2D and 3D measurements.