Abstract：
OBJECTIVE: The purpose of this in vitro study was to compare the trueness and precision of complete arch implant impressions using conventional impression, intraoral scanning with and without splinting, and stereophotogrammetry.
METHODS: An edentulous model with six implants was used in this study. Four implant impression techniques were compared: the conventional impression (CI), intraoral scanning (IOS) without splinting, intraoral scanning with splinting (MIOS), and stereophotogrammetry (SPG). An industrial blue light scanner was used to generate the baseline scan from the model. The CI was captured with a laboratory scanner. The reference best-fit method was then applied in the computer-aided design (CAD) software to compute the three-dimensional, angular, and linear discrepancies among the four impression techniques. The root mean square (RMS) 3D discrepancies in trueness and precision between the four impression groups were analyzed with a Kruskal-Wallis test. Trueness and precision between single analogs were assessed using generalized estimating equations.
RESULTS: Significant differences in the overall trueness (p = .017) and precision (p < .001) were observed across four impression groups. The SPG group exhibited significantly smaller RMS 3D deviations than the CI, IOS, and MIOS groups (p < .05), with no significant difference detected among the latter three groups (p > .05).
CONCLUSIONS: Stereophotogrammetry showed superior trueness and precision, meeting misfit thresholds for implant-supported complete arch prostheses. Intraoral scanning, while accurate like conventional impressions, exhibited cross-arch angular and linear deviations. Adding a splint to the scan body did not improve intraoral scanning accuracy.
METHODS: An edentulous model with six implants was used in this study. Four implant impression techniques were compared: the conventional impression (CI), intraoral scanning (IOS) without splinting, intraoral scanning with splinting (MIOS), and stereophotogrammetry (SPG). An industrial blue light scanner was used to generate the baseline scan from the model. The CI was captured with a laboratory scanner. The reference best-fit method was then applied in the computer-aided design (CAD) software to compute the three-dimensional, angular, and linear discrepancies among the four impression techniques. The root mean square (RMS) 3D discrepancies in trueness and precision between the four impression groups were analyzed with a Kruskal-Wallis test. Trueness and precision between single analogs were assessed using generalized estimating equations.
RESULTS: Significant differences in the overall trueness (p = .017) and precision (p < .001) were observed across four impression groups. The SPG group exhibited significantly smaller RMS 3D deviations than the CI, IOS, and MIOS groups (p < .05), with no significant difference detected among the latter three groups (p > .05).
CONCLUSIONS: Stereophotogrammetry showed superior trueness and precision, meeting misfit thresholds for implant-supported complete arch prostheses. Intraoral scanning, while accurate like conventional impressions, exhibited cross-arch angular and linear deviations. Adding a splint to the scan body did not improve intraoral scanning accuracy.
摘要：
目的：这项体外研究的目的是比较使用常规印模的完整足弓植入物印模的真实性和准确性，有或没有夹板的口内扫描，和立体摄影测量.
方法：本研究使用了六个植入物的无牙模型。比较了四种植入物印模技术：常规印模（CI），没有夹板的口内扫描（IOS），口腔内扫描与夹板（MIOS），和立体摄影测量（SPG）。使用工业蓝光扫描仪从模型生成基线扫描。TheCI是用实验室扫描仪捕获的。然后将参考最佳拟合方法应用于计算机辅助设计(CAD)软件中,角度,四种印象技术之间的线性差异。使用Kruskal-Wallis检验分析了四个印象组之间的均方根（RMS）3D准确性和准确性差异。使用广义估计方程评估单个类似物之间的真实性和准确性。
结果：在四个印象组中观察到总体真实性（p=.017）和精确性（p<.001）的显着差异。SPG组的RMS3D偏差明显小于CI、IOS、和MIOS组（p<0.05），后三组间无显著差异(p>.05)。
结论：立体摄影测量显示出较高的真实性和精确性，满足植入物支持的完整足弓假体的失配阈值。口内扫描,虽然像传统的印象一样准确，表现出交叉拱角度和线性偏差。在扫描主体上添加夹板并不能提高口内扫描的准确性。
方法：本研究使用了六个植入物的无牙模型。比较了四种植入物印模技术：常规印模（CI），没有夹板的口内扫描（IOS），口腔内扫描与夹板（MIOS），和立体摄影测量（SPG）。使用工业蓝光扫描仪从模型生成基线扫描。TheCI是用实验室扫描仪捕获的。然后将参考最佳拟合方法应用于计算机辅助设计(CAD)软件中,角度,四种印象技术之间的线性差异。使用Kruskal-Wallis检验分析了四个印象组之间的均方根（RMS）3D准确性和准确性差异。使用广义估计方程评估单个类似物之间的真实性和准确性。
结果：在四个印象组中观察到总体真实性（p=.017）和精确性（p<.001）的显着差异。SPG组的RMS3D偏差明显小于CI、IOS、和MIOS组（p<0.05），后三组间无显著差异(p>.05)。
结论：立体摄影测量显示出较高的真实性和精确性，满足植入物支持的完整足弓假体的失配阈值。口内扫描,虽然像传统的印象一样准确，表现出交叉拱角度和线性偏差。在扫描主体上添加夹板并不能提高口内扫描的准确性。
