Reproducibility of Three-Dimensional Volumetric Measurement of Periocular Tumor Models.

Abstract

In this study, we measured the volume of customized tumor models in the periocular area using three-dimensional (3D) stereophotogrammetry and evaluated the reproducibility of these measurements. Five tumor models of different colors and sizes were placed in different periocular positions, and 3D facial images were obtained from 68 healthy adult volunteers. Subsequently, the volumes of the tumor models were measured, and the intra- and interrater reproducibility was assessed. The gray 6mm model revealed the highest reliable measurements in both Caucasians (intra- and interrater intraclass correlation coefficients of 0.981 and 0.899, mean absolute difference of 1.446 and 3.327mm3, relative error measurement of 3.497% and 8.120%, technical error of measurement of 1.450 and 3.105mm3, and relative technical error of measurement of 3.506% and 7.580%) and Asians (0.968 and 0.844, 1.974 and 4.067mm3, 4.772% and 9.526%, 2.100 and 4.302mm3, and 5.076% and 10.076%, respectively). The highest reliability of measurements in the lateral upper eyelid (0.88 and 0.95, 4.042 and 3.626mm3, 9.730% and 9.020%, 5.714 and 3.358mm3, and 9.730% and 8.350%, respectively) and medial upper eyelid (0.81 and 0.89, 4.313 and 4.226mm3, 9.730% and 9.020%, 6.098 and 4.069mm3, and 9.730% and 8.350%, respectively) with eyes closed was evident in Caucasians, while the same trend (0.841 and 0.815, 2.828 and 3.757mm3, 9.860% and 9.840%, 4.052 and 4.308mm3, and 9.860% and 9.740%, respectively) was observed in Asians in the medial canthus with eyes closed. This study confirms, for the first time, the high reliability of periocular tumor volume measurements using 3D stereophotogrammetry, suggesting its feasibility for eyelid tumor measurement. Further trials are required to investigate its clinical use for documentation and follow-up of different eyelid tumors.