A novel standardized approach for the 3D evaluation of upper eyelid area and volume.

Background: Three-dimensional (3D) surface imaging and morphometrics are becoming the preferred craniofacial surface imaging modality. However, as a unique advantage of 3D imaging, areal and volumetric measurements have been rarely conducted and validated for evaluating soft tissue change in the periocular region, especially the upper eyelids. Therefore, based on an existing periocular landmark identification strategy, we proposed a novel modified method to define a standardized upper eyelid region for areal and volumetric measurements and validate its reliability for future clinical application.

Methods: Forty-four healthy adult volunteers were enrolled in this cross-sectional study. Three-dimensional facial images were taken with a 3D imaging system. Subsequently, the upper eyelid region selection and areal and volumetric measurements were conducted using a modified landmarks localization strategy to evaluate their intrarater, interrater, and intramethod reliability.

Results: Areal measurement of the upper eyelid revealed highly reliable outcomes for intraclass correlation coefficient (ICC), mean absolute difference (MAD), relative error measurement (REM), technical error of measurement (TEM), and relative technical error of measurement (%TEM) of 0.982, 0.1620 cm, 2.9%, 0.1510 cm, and 2.7% for intrarater reliability, respectively; 0.969, 0.2076 cm, 3.7%, 0.1930 cm, and 3.5% for interrater reliability, respectively; and 0.917, 0.3636 cm, 6.5%, 0.3354 cm, and 6.0% for intramethod reliability, respectively. Unsatisfactory results were found for the volumetric measurement of the upper eyelid: the ICC, MAD, REM, TEM, and %TEM estimates for intrarater reliability were 0.992, 0.2299 mL, 10.3%, 0.2414 mL, and 10.8%, respectively; for interrater reliability, these values were 0.985, 0.2749 mL, 12.3%, 0.3253 mL, and 14.6%, respectively; and for intramethod reliability, these values were 0.433, 1.6716 mL, 77.9%, 2.0615 mL, and 96.1%, respectively.

Conclusions: This is the first study to propose a standardized upper eyelid region selection strategy and simultaneously validate its reliability for 3D areal and volumetric measurements. This study confirmed the high-level reliability of areal measurement and poor reliability of volumetric measurement based on direct measurements using a single image, which may provide better results when this method is combined with the image overlapping and registration procedure. However, this is subject to further validation. Nonetheless, this method could provide quantitative areal and volumetric data on the upper eyelids and might have widespread application potential in the future.