目的：探索一种光学扫描牙冠三维模型和锥形束CT（cone beam CT，CBCT）重建牙根三维模型的数据融合方法，实现不同来源冠、根模型的自然过渡融合。方法： 选取临床口腔正畸下颌牙列轻度拥挤患者1例，应用Mimics 17.0软件对患者CBCT数据重建带有牙根的牙列三维模型，并在Geomagic Stuido 2012软件中与患者高精度解剖形态的光学扫描牙列模型进行模型配准。对配准后的三维模型进行精确冠、龈边界线提取，边界线根向偏置及投影，冠、根边界线生成及模型剪裁等操作，并通过曲率连续算法实现不同来源冠、根模型的自然过渡缝合，完成冠、根三维融合模型的构建。结果：本研究基于商业化的软件平台，初步实现了针对光学扫描三维牙冠数据与CBCT三维牙根数据的曲率移形过渡融合方法，应用该方法完成了正畸临床10例牙列轻度拥挤患者冠、根三维融合模型的构建，融合模型由高年资医师主观打分评价，平均满意度8.6分（0~10分），初步验证了该方法的可行性和有效性。结论： 本研究的冠、根三维数据融合方法可获得逼近真实生理解剖形态的完整牙齿及牙列三维模型，其对复杂牙列拥挤及错颌畸形患者的适应性有待进一步研究。
Objective：To explore a three-dimensional (3D) data fusion and integration method of optical scanning tooth crowns and cone beam CT (CBCT) reconstructing tooth roots for their natural transition in the 3D profile. Methods: One mild dental crowding case was chosen from orthodontics clinics with full denture. The CBCT data were acquired to reconstruct the dental model with tooth roots by Mimics 17.0 medical imaging software, and the optical impression was taken to obtain the dentition model with high precision physiological contour of crowns by Smart Optics dental scanner. The two models were doing 3D registration based on their common part of the crowns’ shape in Geomagic Studio 2012 reverse engineering software. The model coordinate system was established by defining the occlusal plane. crown-gingiva boundary was extracted from optical scanning model manually, then crown-root boundary was generated by offsetting and projecting crown-gingiva boundary to the root model. After trimming the crown and root models, the 3D fusion model with physiological contour crown and nature root was formed by curvature continuity filling algorithm finally. In the study, 10 patients with dentition mild crowded from the oral clinics were followed up with this method to obtain 3D crown and root fusion models, and 10 high qualification doctors were invited to do subjective evaluation of these fusion models. Results: This study based on commercial software platform, preliminarily realized the 3D data fusion and integration method of optical scanning tooth crowns and CBCT tooth roots with a curvature continuous shape transition. The 10 patients’ 3D crown and root fusion models were constructed successfully by the method, and the average score of the doctors’ subjective evaluation for these 10 models was 8.6 points (0-10 points). which meant that all the fusion models could basically meet the need of the oral clinics, and also showed the method in our study was feasible and efficient in orthodontics study and clinics. Conclusion: The method of this study for 3D crown and root data fusion could obtain an integrate tooth or dental model more close to the nature shape. CBCT model calibration may probably improve the precision of the fusion model. The adaptation of this method for severe dentition crowding and micromaxillary deformity needs further research.
