Objective: To comparatively evaluate the accuracy and the scan time of three full-arch scan strategies on the head-simulator, to explore a full-arch scan strategy with better clinical operability and high accuracy. Methods: A cross-controlled study design was used. A model with melamine-formaldehyde resin teeth and silica gel gingiva of an upper dental arch which can be fixed on a head simulator was scanned with an optical scanner (ATOS Core) in order to obtain the standard tessellation language (STL) dataset as reference. Intraoral scans were performed on the model fixed on the head simulator with four intraoral scanners (IOS) [A (TRIOS 3), B (CS 3600), C (CEREC Omnicam), D (iTero)]. The STL datasets were obtained from each of the four different IOS systems by using three scan strategies (scan strategies 1, 2 and 3 were composed of 10, 5 and 7 paths respectively) all by one attending doctor with 3 years of intraoral scanning experience. For each scanner and each scan strategy, nine scans were acquired. And the scan time was recorded for each scan. Following the scan strategy, the scan path was completed to obtain a full-arch digital model, and the scan time was recorded as full-arch scan time. Complementary scans were performed to fill the missing image, and this scan time was recorded as complementary scan time. The total scan time was obtained by adding full-arch scan time and complementary scan time. Through the Geomagic Wrap software, the three-dimensional (3D) models were overlaid by best fit alignment function and compared to obtain the root mean square values of the discrepancies by 3D compare function. The intraoral scanning datasets were compared with the reference for trueness. The nine intraoral scanning datasets were cross compared with same scan strategy and same intraoral scanner for precision. Results: There were no significant differences among the three scan strategies for trueness (P>0.05), while the differences among the three scan strategies for precision were affected by difference IOSs (P<0.05), and only scan strategy 3 showed the highest precision with all the four IOS. The full-arch scan time of scan strategies 1, 2 and 3 were (130±24), (72±17) and (90±19) s respectively (P<0.05). For complementary scan time, scan strategy 2 [(50±24) s] took longer time than scan strategy 1 [(26±18) s] and scan strategy [(25±21) s] (P<0.05), while no significant differences between the latter two (P>0.05). For total scan time, scan strategy 1 [(156±31) s] took longer time than scan strategy 2 [(122±30) s ] and scan strategy 3 [(115±29) s ] (P<0.05), while no significant differences between the latter two (P>0.05). Conclusions: Full-arch scanning on the head-simulator with scan strategy 3 which can obtain scanning datasets with high accuracy, was more convenient to operate and took shorter scan time, and is generally suitable for intraoral scanners commonly used in clinic.
目的： 对比评价仿真头颅模型环境下口内扫描仪3种全牙弓扫描策略的扫描精度和扫描时间，探讨临床适应性更强且可获得高扫描精度的全牙弓扫描策略。 方法： 采用相互对照实验设计，使用光学扫描仪ATOS Core扫描上颌模型，获取“.stl”格式数据作为参考数据。将上颌模型固定于仿真头颅模型上，由1名具有3年口内扫描经验的主治医师使用4种口内扫描仪［A（TRIOS 3）、B（CS 3600）、C（CEREC Omnicam）、D（iTero）］，分别使用3种扫描策略（扫描策略1、2、3分别由10、5、7段路径组成）各扫描9次，获取“.stl”格式数据作为口扫数据，并记录扫描时间：完成所有扫描路径、获得1个完整的全牙弓数字化模型的时间为全牙弓扫描时间；补扫模型表面图像空缺处的时间为补扫时间；全牙弓扫描时间与补扫时间之和为总扫描时间。使用Geomagic Wrap软件的最佳拟合功能，获得口扫数据与参考数据三维偏差的均方根值，以评价扫描正确度；对同一口内扫描仪同一扫描策略9次扫描获得的口扫数据进行两两比较，获得均方根值以评价扫描精密度。 结果： 使用同一扫描仪的条件下，3种扫描策略扫描正确度的总体差异均无统计学意义（P>0.05）；3种扫描策略扫描精密度的总体差异均有统计学意义（P<0.05），其中扫描策略3在各扫描仪中的扫描精密度均为最高。扫描策略1、2、3的全牙弓扫描的时间分别为（130±24）、（72±17）和（90±19）s，各组间差异均有统计学意义（P<0.05）。扫描策略1和3的补扫时间分别为（26±18）和（25±21）s（P>0.05），均显著少于扫描策略2［（50±24）s］（P<0.05）。扫描策略2和3的总扫描时间［（122±30）和（115±29）s（P>0.05）］显著少于扫描策略1［（156±31）s］（P<0.05）。 结论： 仿真头颅模型环境下使用扫描策略3进行全牙弓扫描，可获得高精度扫描数据，操作更便捷，扫描时间更短，且普适于目前临床常用的口内扫描仪。.