Abstract：
The objective of this study was to investigate the accuracies of three-dimensional (3D) facial scanning mobile phone applications as compared to professional 3D facial imaging systems.
A manikin head model was used as the subject for comparing six 3D facial imaging systems which comprised three professional 3D scanners (3dMDface, Artec Eva and Vectra H2) and three mobile phone applications (Bellus3D, ScandyPro and Hedges). For each system, five scans were taken to analyse (1) linear accuracy using 9 measurements (2) global and (3) regional 3D accuracy of the scanned surface by root mean square (RMS) and colour map analysis. Another set of five scans was repeated by a second operator to evaluate the inter-operator reproducibility for each system.
All the facial imaging systems had absolute errors lesser than 1.0 mm for the linear measurements. The technical error of measurement (TEM) for inter-examiner and intra-examiner linear measurements were within acceptable limits. Artec Eva, Vectra H2 and Scandy Pro had poor global 3D trueness (RMS > 1.0 mm) but good 3D regional trueness (RMS < 1.0 mm). 3dMDface, Bellus3D Face App and Heges had good global and regional 3D trueness. All the facial imaging systems had good global and regional 3D precision and reproducibility (RMS < 1.0 mm).
This study demonstrated that mobile phone 3D scanning applications had comparable trueness, precision and reproducibility to professional systems. Colour map analysis supplemented the use of the RMS value to demonstrate facial regions of significant deviation. Clinicians should also consider the specific area or region of inaccuracies for each system to determine whether the chosen system is appropriate for the clinical condition or procedure.
Mobile phone 3D facial imaging applications may be as accurate as 3D professional facial scanning systems for craniomaxillofacial purposes. However, the choice of the system may vary depending on the specific area of interest.
A manikin head model was used as the subject for comparing six 3D facial imaging systems which comprised three professional 3D scanners (3dMDface, Artec Eva and Vectra H2) and three mobile phone applications (Bellus3D, ScandyPro and Hedges). For each system, five scans were taken to analyse (1) linear accuracy using 9 measurements (2) global and (3) regional 3D accuracy of the scanned surface by root mean square (RMS) and colour map analysis. Another set of five scans was repeated by a second operator to evaluate the inter-operator reproducibility for each system.
All the facial imaging systems had absolute errors lesser than 1.0 mm for the linear measurements. The technical error of measurement (TEM) for inter-examiner and intra-examiner linear measurements were within acceptable limits. Artec Eva, Vectra H2 and Scandy Pro had poor global 3D trueness (RMS > 1.0 mm) but good 3D regional trueness (RMS < 1.0 mm). 3dMDface, Bellus3D Face App and Heges had good global and regional 3D trueness. All the facial imaging systems had good global and regional 3D precision and reproducibility (RMS < 1.0 mm).
This study demonstrated that mobile phone 3D scanning applications had comparable trueness, precision and reproducibility to professional systems. Colour map analysis supplemented the use of the RMS value to demonstrate facial regions of significant deviation. Clinicians should also consider the specific area or region of inaccuracies for each system to determine whether the chosen system is appropriate for the clinical condition or procedure.
Mobile phone 3D facial imaging applications may be as accurate as 3D professional facial scanning systems for craniomaxillofacial purposes. However, the choice of the system may vary depending on the specific area of interest.
摘要：
背景：三维（3D）面部扫描系统的使用已成为治疗颅颌面疾病的不可或缺的一步。这项研究的目的是调查与专业3D面部成像系统相比，三维（3D）面部扫描手机应用的准确性。
方法：使用人体模型作为比较六个3D面部成像系统的主题，其中包括三个专业3D扫描仪（3dMDface，ArtecEva和VectraH2)和三个手机应用程序(Bellus3D、ScandyPro和对冲）。对于每个系统，进行了五次扫描以分析（1）使用9次线性测量的线性精度（2）通过均方根（RMS）和彩色图分析扫描表面的全局和（3）区域3D精度。由第二操作者重复另一组五次扫描以评估每个系统的操作者间再现性。
结果：对于线性测量，所有面部成像系统的绝对误差均小于1.0mm。检查者之间和检查者内部线性测量的技术测量误差（TEM）在可接受的范围内。ArtecEva,VectraH2和ScandyPro的整体3D真实性较差（RMS>1.0mm），但3D区域真实性较好（RMS<1.0mm）。3dMDface,Bellus3DFaceApp和Heges具有良好的全球和区域3D真实性。所有面部成像系统均具有良好的全局和区域3D精度和3D再现性（RMS<1.0mm）。
结论：这项研究表明，手机3D扫描应用具有相当的真实性，专业系统的精度和再现性。色图分析补充了RMS值的使用，以证明明显偏差的面部区域。临床医生还应考虑每个系统的特定区域或不准确区域，以确定所选择的系统是否适合临床状况或程序。
结论：更简单的手机3D面部成像应用可能与用于颅颌面目的的3D专业面部扫描系统一样准确。然而,系统的选择可以根据感兴趣的特定区域而变化。
方法：使用人体模型作为比较六个3D面部成像系统的主题，其中包括三个专业3D扫描仪（3dMDface，ArtecEva和VectraH2)和三个手机应用程序(Bellus3D、ScandyPro和对冲）。对于每个系统，进行了五次扫描以分析（1）使用9次线性测量的线性精度（2）通过均方根（RMS）和彩色图分析扫描表面的全局和（3）区域3D精度。由第二操作者重复另一组五次扫描以评估每个系统的操作者间再现性。
结果：对于线性测量，所有面部成像系统的绝对误差均小于1.0mm。检查者之间和检查者内部线性测量的技术测量误差（TEM）在可接受的范围内。ArtecEva,VectraH2和ScandyPro的整体3D真实性较差（RMS>1.0mm），但3D区域真实性较好（RMS<1.0mm）。3dMDface,Bellus3DFaceApp和Heges具有良好的全球和区域3D真实性。所有面部成像系统均具有良好的全局和区域3D精度和3D再现性（RMS<1.0mm）。
结论：这项研究表明，手机3D扫描应用具有相当的真实性，专业系统的精度和再现性。色图分析补充了RMS值的使用，以证明明显偏差的面部区域。临床医生还应考虑每个系统的特定区域或不准确区域，以确定所选择的系统是否适合临床状况或程序。
结论：更简单的手机3D面部成像应用可能与用于颅颌面目的的3D专业面部扫描系统一样准确。然而,系统的选择可以根据感兴趣的特定区域而变化。
