实用新型的CBCT重建模型与激光扫描牙齿STL模型的融合方法
发布时间：2019-06-27 18:16
【摘要】：目的本研究将患者的CBCT数据和激光扫描的高精度上下颌牙齿模型的数据采用局部配准的方法进行融合,建立数字化三维颅(?)模型,并对配准的精度进行评价。为评价正畸矫治效果尤其是隐形矫治效果及口腔颌面外科获取精确患者咬合模型提供简便且精确性可评估的方法。材料和方法本研究选取 2015 到 2016 年间接受隐适美(Invisalign,Align Technology,Santa ClaraCA,USA)正畸治疗的10名成年患者,这些患者的所有恒牙均已经萌出,不存在其他明显的颅面结构异常。所有患者由同一操作者进行NewTom 5G(QR,Verona,Italy)CBCT扫描,扫描数据以DICOM格式输出并储存。我们将数据导入到 Mimics17.0(Materialise's interactive medical image control system)软件中,运用Mimics进行阈值选择,形成上下颌颌骨及各自牙列相对独立且保有原始相对位置的3D数字化模型。利用两步法硅橡胶取模技术,取得患者上下颌牙列及周围软组织的精确阴模,用牙科超硬石膏灌制模型,从而得到精确的石膏阳模。利用Activity 888高精度扫描仪(Smart OpticsGnbH,Bochum,Germany)将石膏阳模数字化,以STL文件格式导出。将激光扫描得到的STL文件导入Mimics软件,利用Mimics软件的点配准功能及STL配准功能,先选取牙齿(?)面较有特征的结构进行点配准,之后对激光扫描产生的STL数据与重建的CBCT模型以牙列的所有临床冠为基准进行局部配准。配准完成后,将配准结果以MCS文件格式保存。利用3-matic9.0软件打开MCS文件。首先在从扫描仪获取的牙齿模型上选出牙列的临床冠部分。接着通过计算CBCT模型和扫描仪获取的STL模型的空间相对的点云中的点到点的距离可得到两模型在该区域距离的均方根值(RootMeanSquare,RMS)即综合平均距离。当RMS小于0.3mm(CBCT图像体素值)时,可认为两个模型的配准精度好,误差可忽略。配准精度得到确认之后,在Mimics中打开之前保存的MCS文件,将CBCT模型的临床牙冠部分去除,融合CBCT模型与STL牙列模型。结果1.距离云图的分析从距离云图上可以看出,大部分的临床牙冠区域的颜色为绿色,说明两个模型在该区域的距离趋近于0。2.CBCT重建模型与扫描模型在临床冠部分的配准后距离分析利用软件提供的数据,可计算出两平面的RMS数值均小于0.3mm。3.上下颌牙列咬合情况的三维重建完成模型后,咬合情况可准确再现。结论1.通过计算两模型综合相对距离(RMS)值的方式,量化地体现和评估了数字化激光扫描模型和CBCT重建的模型局部配准后的精度。2.利用扫描获得的高精度牙列数字模型与CBCT重建模型的局部配准可获得具有高清晰度牙列,咬合关系,龈缘形态,头颅骨骼结构的3D数字化模型。
[Abstract]:Objective in this study, the CBCT data of the patients and the data of the high precision maxillary and mandibular teeth model scanned by laser were combined with the local registration method, and the digital three-dimensional skull (?) Model, and the accuracy of registration is evaluated. It provides a simple and accurate method for evaluating the effect of orthodontic correction, especially invisible correction, and obtaining accurate occlusal model in oral and maxillofacial surgery. Materials and methods Ten adult patients undergoing orthodontic treatment with Invisalign,Align Technology,Santa ClaraCA,USA from 2015 to 2016 were selected. All the permanent teeth of these patients had sprouted and there were no other obvious craniofacial structural abnormalities. All patients were scanned with NewTom 5G (QR,Verona,Italy) CBCT by the same operator, and the scanning data were output and stored in DICOM format. We import the data into Mimics17.0 (Materialise's interactive medical image control system) software and use Mimics for threshold selection to form a 3D digital model of jaw and their dentition which is relatively independent and maintains the original relative position. By using two-step silicone rubber molding technique, the accurate negative mold of maxillary dentition and surrounding soft tissue was obtained, and the dental superanhydrite filling model was used to obtain the accurate plaster positive mold. The gypsum positive mould is digitized by Activity 888 high precision scanner (Smart OpticsGnbH,Bochum,Germany) and exported in STL file format. The STL file obtained by laser scanning is imported into Mimics software. By using the point registration function and STL registration function of Mimics software, the teeth (?) The point registration of the structure with characteristic surface is carried out, and then the STL data generated by laser scanning and the reconstructed CBCT model are locally registered on the basis of all the clinical crowns of the dentition. After registration, the registration results are saved in MCS file format. Use 3-matic9.0 software to open MCS file. First, the clinical crown part of the dentition was selected on the tooth model obtained from the scanner. Then the root mean square value (RootMeanSquare,RMS) of the distance between the two models in this region can be obtained by calculating the distance from point to point in the space relative to the point cloud of the STL model obtained by the CBCT model and the scanner. When RMS is less than 0.3mm (CBCT image voxel value), it can be considered that the registration accuracy of the two models is good and the error can be ignored. After the registration accuracy is confirmed, the MCS file saved before is opened in Mimics, the clinical crown part of CBCT model is removed, and the CBCT model and STL dentition model are combined. Result 1. The analysis of distance cloud image shows that the color of most clinical crown areas is green, which indicates that the distance between the two models in this area is closer to that of 0.2.CBCT reconstruction model and scanning model after registration of clinical crown part. Using the data provided by the software, it can be calculated that the RMS values of both planes are less than 0.3 mm. 3. After the three-dimensional reconstruction of maxillary and maxillary dentition occlusal condition is completed, the occlusal situation can be reproduced accurately. Conclusion 1. By calculating the relative distance (RMS) value of the two models, the accuracy of local registration of digital laser scanning model and CBCT reconstructed model is quantitatively reflected and evaluated. 2. 3D digital model with high definition dentition, occlusal relationship, gum edge shape and skull bone structure can be obtained by local registration of high precision dentition digital model and CBCT reconstruction model obtained by scanning.
【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R783.5
本文编号：2507026
[Abstract]:Objective in this study, the CBCT data of the patients and the data of the high precision maxillary and mandibular teeth model scanned by laser were combined with the local registration method, and the digital three-dimensional skull (?) Model, and the accuracy of registration is evaluated. It provides a simple and accurate method for evaluating the effect of orthodontic correction, especially invisible correction, and obtaining accurate occlusal model in oral and maxillofacial surgery. Materials and methods Ten adult patients undergoing orthodontic treatment with Invisalign,Align Technology,Santa ClaraCA,USA from 2015 to 2016 were selected. All the permanent teeth of these patients had sprouted and there were no other obvious craniofacial structural abnormalities. All patients were scanned with NewTom 5G (QR,Verona,Italy) CBCT by the same operator, and the scanning data were output and stored in DICOM format. We import the data into Mimics17.0 (Materialise's interactive medical image control system) software and use Mimics for threshold selection to form a 3D digital model of jaw and their dentition which is relatively independent and maintains the original relative position. By using two-step silicone rubber molding technique, the accurate negative mold of maxillary dentition and surrounding soft tissue was obtained, and the dental superanhydrite filling model was used to obtain the accurate plaster positive mold. The gypsum positive mould is digitized by Activity 888 high precision scanner (Smart OpticsGnbH,Bochum,Germany) and exported in STL file format. The STL file obtained by laser scanning is imported into Mimics software. By using the point registration function and STL registration function of Mimics software, the teeth (?) The point registration of the structure with characteristic surface is carried out, and then the STL data generated by laser scanning and the reconstructed CBCT model are locally registered on the basis of all the clinical crowns of the dentition. After registration, the registration results are saved in MCS file format. Use 3-matic9.0 software to open MCS file. First, the clinical crown part of the dentition was selected on the tooth model obtained from the scanner. Then the root mean square value (RootMeanSquare,RMS) of the distance between the two models in this region can be obtained by calculating the distance from point to point in the space relative to the point cloud of the STL model obtained by the CBCT model and the scanner. When RMS is less than 0.3mm (CBCT image voxel value), it can be considered that the registration accuracy of the two models is good and the error can be ignored. After the registration accuracy is confirmed, the MCS file saved before is opened in Mimics, the clinical crown part of CBCT model is removed, and the CBCT model and STL dentition model are combined. Result 1. The analysis of distance cloud image shows that the color of most clinical crown areas is green, which indicates that the distance between the two models in this area is closer to that of 0.2.CBCT reconstruction model and scanning model after registration of clinical crown part. Using the data provided by the software, it can be calculated that the RMS values of both planes are less than 0.3 mm. 3. After the three-dimensional reconstruction of maxillary and maxillary dentition occlusal condition is completed, the occlusal situation can be reproduced accurately. Conclusion 1. By calculating the relative distance (RMS) value of the two models, the accuracy of local registration of digital laser scanning model and CBCT reconstructed model is quantitatively reflected and evaluated. 2. 3D digital model with high definition dentition, occlusal relationship, gum edge shape and skull bone structure can be obtained by local registration of high precision dentition digital model and CBCT reconstruction model obtained by scanning.
【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R783.5
【引证文献】
相关硕士学位论文 前1条
1 张润生;数字化口腔种植导板设计及3D打印关键问题研究[D];河北科技大学;2018年
，本文编号：2507026
本文链接：https://www.wllwen.com/yixuelunwen/kouq/2507026.html
最近更新
教材专著
