目的

通过比较三维数字模型测量和传统石膏模型测量的差异，探讨三维数字模型测量在美国正畸专家认证委员会（ABO）客观评分系统（OGS）应用的可重复性和准确性。

方法

选择常规正畸治疗后的石膏模型30 副，采用3Shape D810 激光扫描仪扫描生成数字化模型，用计算机软件进行三维数字模型测量；原石膏模型经修整制作成标准石膏模型，用于传统石膏模型测量。 测量项目包括ABO 客观评分系统中的7 项标准，即牙齿排列、边缘嵴、颊舌向倾斜、咬合接触、咬合关系、覆盖和邻接关系。根据测量结果计算每项标准的得分及总得分。所有测量均由第一作者本人完成，先进行数字模型测量，2 周后重复测量1 次，数字测量结束2 周后进行石膏模型测量，同样于2 周后重复测量1 次。 用Wilcoxon 法检验和比较两种方法的可重复性，并比较两种方法的测量结果。

结果

对于两种方法的可重复性检测结果显示，ABO-OGS 总得分及各项标准的得分在两次测量结果中差异均无统计学意义（P＞0.05）；对两种方法可重复性的比较发现，三维数字模型测量在牙齿排列（Z=-3.158，P=0.002）、颊舌向倾斜（Z=-2.555，P=0.011）和咬合接触（Z=-3.150，P=0.002）三项标准的得分及总得分（Z=-2.217，P=0.027）的差值小于传统石膏模型测量；比较两种方法的测量结果发现，三维数字模型测量在牙齿排列（Z=-3.105，P=0.002）、咬合接触（Z=-2.415，P=0.016）、覆盖（Z=-2.818，P=0.005）三项标准的得分和总得分（Z=-4.149，P＜0.001）高于传统石膏模型测量。

结论

在ABO 客观评分系统应用中，三维数字模型测量比传统石膏模型测量具有更好的可重复性和准确性。

Objective

To evaluate the reproducibility and accuracy of 3D digital model measurements for the American Board of Orthodontics objective grading system by comparing with the traditional plaster model measurements.

Methods

Thirty conventional orthodontic posttreatment plaster models were included in this study. The plaster models were scanned by 3Shape D810 laser scanner to produce corresponding digital models for 3D digital model measurements with computer software. Then they were trimmed to be standard plaster models for traditional plaster model measurements. Seven criteria （tooth alignment，marginal ridges，buccolingual inclination，occlusal contacts，occlusal relationship，overjet， and interproximal contacts） of the ABO objective grading system were measured. All the measurements were performed by the author. Digital model measurements were performed twice at 2 weeks interval. When finished after 2 weeks， plaster model measurements were also performed twice at 2 weeks interval. After the completion of the measurement，each criterion score and total score were calculated and the results were recorded. Wilcoxon test was used to analyz the easurement results and compare the reproducibility of two methods.

Results

For the reproducibility of two methods， the ABOOGS total score and each criterion score were no significance in two measurements performed at different time （P＞0.05）. Comparing the reproducibility of the two methods， 3D digital model measurements showed lower score difference in tooth alignment（Z=-3.158，P=0.002）， buccolingual inclination（Z=-2.555， P=0.011）， occlusal contacts （Z=-3.150， P=0.002） and total （Z=-2.217， P=0.027）. By comparison of the measurements of two methods， 3D digital model measurements showed higher score in tooth alignment （Z=-3.105，P=0.002）， occlusal contacts （Z=-2.415，P=0.016）， overjet（Z=-2.818，P=0.005） and total （Z=-4.149，P＜0.001）.

Conclusions

Three-dimensional digital model measurements have better reproducibility and accuracy for the American Board of Orthodontics objective grading system than traditional plaster model measurements.