Private Practice, 00198 Rome, Italy.
Department of Industrial Engineering, University of Rome "Tor Vergata", 00133 Rome, Italy.
J Dent. 2022 Jan;116:103881. doi: 10.1016/j.jdent.2021.103881. Epub 2021 Nov 8.
The 3D facial scan technology allows to virtualize the face of the patient, that can be incorporated with other 3D dental images produced by digital scanning of the dental structures. Aim of this study is to investigate the trueness and precision of a low-cost portable face scanner, with two different scan techniques MATERIALS AND METHODS: Ten patients were enrolled for this study and seventeen soft tissue landmarks were selected to perform linear facial measurements, specifically Reference (Ref), Pronasion (Prn), Subnasal (Sn), Exocanthion Left (Ex-L), Exocanthion Right (Ex-R), Pogonion (Pg), Glabella (G), Alar curvature Right (Al-R), Alar curvature Left (Al-L), Zygion Left (Zn-L), Zygion Right (Zn-R), Orbital Left (Or-L), Orbital Right (Or-R), Tragus Right (T-R), Tragus Left (T-L), Chelion Right (Ch-R) and Chelion Left (Ch-L). Interlandmark distances were measured both manually and digitally. For the manual group ten measurements were made using a digital caliper. For digital group measurements were recorded on the patient face scan obtained using an Ipad Pro 3rd Gen. (Apple Store, Cupertino, CA, USA) and Bellus3D Dental Pro-App (Bellus3D, Inc. Campbell, CA, USA) using "face mode" scan with two different scanning techniques, named Free technique (FT) and Slider Technique (ST). Ten measurements were made for each technique. An open-source software (Meshlab; Meshlab) was used to record all the distances. A paired t-test was used to analyze FT and ST results. In order to further evaluate precision and scan repeatability a surface analysis was performed with both scanning techniques using a CAD software (GOM inspect, GOM) and the total differences in absolute 3D deviations were calculated as root mean square.
The comparison between manual and digital measurements showed a mean absolute difference of 0.95±0.25 for FT and 1.00±0.29 for the ST. Trueness analysis showed statistically significant differences for the Exocanthion L- Exocanthion R measurement with FT having better performance (P<.05). Precision analysis showed statistically significant differences for G-Pg, Ref-Zn-R and Prn-Zn-R with ST having better performance (P<.05). To achieve all the scans required without any signs of deformation, 184 scans were performed using Free technique and 124 scans using Slider technique. Surface analysis revealed a mean distance of 0.12±0.45 between Free scans and 0.13±0.46 between Slider scans in accordance with the linear measurement analysis CONCLUSION: The study showed that accuracy of low-cost portable scanner can be suitable for clinical use. The use of ST is suggested for a reliable clinical use due to the better precision and an effective reduction of motion artifacts and the lower compliance required to the patients during the scan.
3D 面部扫描技术可使患者的面部实现虚拟化,且可与通过数字扫描牙齿结构获得的其他 3D 牙科图像结合使用。本研究旨在调查一种低成本便携式面部扫描仪的准确性和精密度,该扫描仪使用两种不同的扫描技术。
本研究纳入了 10 名患者,并选择了 17 个软组织标志点进行线性面部测量,具体包括参考点(Ref)、前突点(Prn)、鼻下点(Sn)、外眦点左侧(Ex-L)、外眦点右侧(Ex-R)、颏前点(Pg)、眉间点(G)、右侧鼻翼侧拐点(Al-R)、左侧鼻翼侧拐点(Al-L)、颧突点左侧(Zn-L)、颧突点右侧(Zn-R)、眶点左侧(Or-L)、眶点右侧(Or-R)、耳屏点右侧(T-R)、耳屏点左侧(T-L)、唇珠点右侧(Ch-R)和唇珠点左侧(Ch-L)。使用手动和数字两种方式分别测量各标志点之间的距离。手动组使用数字卡尺进行了 10 次测量。数字组则使用 iPad Pro 3 代(Apple Store,Cupertino,CA,USA)和 Bellus3D Dental Pro-App(Bellus3D,Inc.,Campbell,CA,USA)上的“面部模式”扫描,使用两种不同的扫描技术(分别命名为自由技术(FT)和滑动技术(ST))获得患者面部扫描图像。每种技术都进行了 10 次测量。使用开源软件(Meshlab;Meshlab)记录所有距离。使用配对 t 检验分析 FT 和 ST 的结果。为了进一步评估精度和扫描重复性,使用 CAD 软件(GOM inspect,GOM)对面部两种扫描技术进行了表面分析,并计算了绝对 3D 偏差的总差异作为均方根。
手动和数字测量结果的比较显示,FT 的平均绝对差异为 0.95±0.25,ST 的平均绝对差异为 1.00±0.29。准确性分析显示,FT 在 Exocanthion L-Exocanthion R 测量中的表现优于 ST(P<.05)。精度分析显示,ST 在 G-Pg、Ref-Zn-R 和 Prn-Zn-R 测量中的表现优于 FT(P<.05)。使用自由技术完成所有扫描共需 184 次,使用滑动技术则需 124 次。表面分析显示,自由扫描之间的平均距离为 0.12±0.45,滑动扫描之间的平均距离为 0.13±0.46,与线性测量分析结果一致。
本研究表明,低成本便携式扫描仪的准确性可满足临床使用需求。由于 ST 具有更高的精度和更有效的减少运动伪影,以及对患者扫描时的配合度要求更低,因此建议用于可靠的临床应用。