Pennington Biomedical Research Center, Louisiana State University System, Baton Rouge, Louisiana, USA.
Louisiana State University, Baton Rouge, Louisiana, USA.
Am J Phys Anthropol. 2021 Aug;175(4):865-875. doi: 10.1002/ajpa.24243. Epub 2021 Feb 5.
Body surface area (SA) is a widely used physical measure incorporated into multiple thermophysiology and evolutionary biology models currently estimated in humans either with empirical prediction equations or costly whole-body laser imaging systems. The introduction of low-cost 3D scanners provides a new opportunity to quantify total body (TB) and regional SA, although a critical question prevails: can these devices acquire the quality of depth information and process this initial data to form a mesh that has the fidelity needed to generate accurate SA estimates?
This question was answered by comparing SA estimates calculated using images from four commercial 3D scanners in 108 adults to corresponding estimates acquired with a whole-body laser system. This was accomplished by processing initial mesh data from all devices, including the laser system, with the same universal software adapted specifically for repairing mesh gaps, identifying landmarks, and generating SA measurements.
TB SA measured on all four 3D scanners was highly correlated with corresponding laser system estimates (R s, 0.98-0.99; all p < 0.001) with some small but significant mean differences (-0.19 to 0.06 m ); root-mean square errors (RMSEs) were small (0.02-0.03 m ); and significant bias was present for one device. Qualitatively similar results (e.g., R s, 0.78-0.95; mean Δs, -0.05 to 0.02 m ; RMSEs, 0.01-0.03 m ) were present for trunk, arm, and leg SA comparisons.
The current study observations demonstrate that low-cost and practical 3D optical scanners are capable of accurately quantifying TB and regional SA, thus opening new opportunities for evaluating human phenotypes and related physiological characteristics.
体表面积（SA）是一种广泛应用的物理测量指标，目前已被纳入多种热生理学和进化生物学模型中，其在人体中的估算方法既有经验预测方程，也有昂贵的全身激光成像系统。低成本的 3D 扫描仪的引入为定量测量全身（TB）和局部 SA 提供了新的机会，尽管存在一个关键问题：这些设备能否获取足够的深度信息，并处理这些初始数据以形成一个具有足够保真度的网格，从而生成准确的 SA 估算值？
为了回答这个问题，我们比较了 108 名成年人的 4 种商用 3D 扫描仪图像计算出的 SA 估算值与全身激光系统获得的相应估算值。这是通过对所有设备（包括激光系统）的初始网格数据进行处理来实现的，包括使用专门针对修复网格间隙、识别地标和生成 SA 测量值的通用软件。
所有 4 种 3D 扫描仪上测量的 TB SA 与相应的激光系统估算值高度相关（R s ，0.98-0.99；所有 p 值均<0.001），存在一些小但有统计学意义的平均差异（-0.19 至 0.06 m ）；均方根误差（RMSE）较小（0.02-0.03 m ）；一台设备存在显著偏差。类似的定性结果（例如，R s ，0.78-0.95；平均差异 Δs ，-0.05 至 0.02 m ；RMSE，0.01-0.03 m ）也存在于躯干、手臂和腿部 SA 比较中。
本研究的观察结果表明，低成本且实用的 3D 光学扫描仪能够准确地定量测量 TB 和局部 SA，从而为评估人类表型和相关生理特征开辟了新的机会。
