Clinical anthropometrics and body composition from 3D whole-body surface scans
Clinical anthropometrics and body composition from 3D whole-body surface scans
复制标题
DOI：
10.1038/ejcn.2016.109
复制
发表时间：
2016-11-01
影响因子：
4.7
通讯作者：
Shepherd, J. A.
中科院分区：
文献类型：
--
作者：
Ng, B. K.;Hinton, B. J.;Shepherd, J. A.
背景/目的：肥胖是一种重要的世界性流行病，需要方便的工具来进行强有力的身体成分分析。我们调查了广泛使用的3D体表扫描仪是否可以提供临床相关的直接人体测量（周长、面积和体积）和身体成分估计（区域脂肪/瘦质量）。受试者/方法：39名按年龄、性别和体重指数（BMI）分层的健康成年人接受了全身3D扫描、双能x线吸收仪（DXA）、空气置换容积脉搏波和磁带测量。进行线性回归来评估三维测量和标准方法之间的一致性。从三维扫描测量中导出线性模型来预测DXA身体成分。37名外部健身中心用户接受了3D扫描和生物电阻抗分析，以验证模型。结果：三维身体扫描测量值与判定方法相关性强：腰围R-2 = 0.95，臀围R-2 = 0.92，表面积R-2 = 0.97，体积R-2 = 0.99。然而，由于地标定位的差异，每个测量结果都存在系统差异。预测体成分方程与整个体（脂肪质量R-2 = 0.95，均方根误差(RMSE) = 2.4 kg）具有较强的一致性；无脂质量R-2 = 0.96, RMSE = 2.2 kg)和手臂、腿和躯干（R-2 = 0.79-0.94, RMSE = 0.5-1.7 kg）。内脏脂肪预测结果一致（R-2 = 0.75, RMSE = 0.11 kg）。结论：三维表面扫描仪提供了精确、稳定的人体形状和成分的自动测量。可能需要软件更新来解决由地标定位差异引起的测量偏差。进一步的研究有理由阐明不同性别、年龄、身体质量指数和种族群体以及代谢障碍患者的体型、身体成分和代谢健康之间的关系。
BACKGROUND/OBJECTIVES: Obesity is a significant worldwide epidemic that necessitates accessible tools for robust body composition analysis. We investigated whether widely available 3D body surface scanners can provide clinically relevant direct anthropometrics (circumferences, areas and volumes) and body composition estimates (regional fat/lean masses).SUBJECTS/METHODS: Thirty-nine healthy adults stratified by age, sex and body mass index (BMI) underwent whole-body 3D scans, dual energy X-ray absorptiometry (DXA), air displacement plethysmography and tape measurements. Linear regressions were performed to assess agreement between 3D measurements and criterion methods. Linear models were derived to predict DXA body composition from 3D scan measurements. Thirty-seven external fitness center users underwent 3D scans and bioelectrical impedance analysis for model validation.RESULTS: 3D body scan measurements correlated strongly to criterion methods: waist circumference R-2 = 0.95, hip circumference R-2 = 0.92, surface area R-2 = 0.97 and volume R-2 = 0.99. However, systematic differences were observed for each measure due to discrepancies in landmark positioning. Predictive body composition equations showed strong agreement for whole body (fat mass R-2 = 0.95, root mean square error (RMSE) = 2.4 kg; fat-free mass R-2 = 0.96, RMSE = 2.2 kg) and arms, legs and trunk (R-2 = 0.79-0.94, RMSE = 0.5-1.7 kg). Visceral fat prediction showed moderate agreement (R-2 = 0.75, RMSE = 0.11 kg).CONCLUSIONS: 3D surface scanners offer precise and stable automated measurements of body shape and composition. Software updates may be needed to resolve measurement biases resulting from landmark positioning discrepancies. Further studies are justified to elucidate relationships between body shape, composition and metabolic health across sex, age, BMI and ethnicity groups, as well as in those with metabolic disorders.