Xin Pengfei, Yu Hongbo, Cheng Huanchong, Shen Shunyao, Shen Steve G F
From the *Department of Oral and Craniomaxillofacial Surgery, Ninth People's Hospital, College of Stomatology, Shanghai Jiao Tong University School of Medicine, Shanghai Key Laboratory of Stomatology, Shanghai 200011; and †Institute of CIM, Shanghai Jiao Tong University, Shanghai Key Laboratory of Advanced Manufacturing Environment, Shanghai 200030, People's Republic of China.
J Craniofac Surg. 2013 Sep;24(5):1573-6. doi: 10.1097/SCS.0b013e3182688ed7.
The aim of this study was to demonstrate the feasibility of building a craniofacial virtual reality model by image fusion of 3-dimensional (3D) CT models and 3 dMD stereophotogrammetric facial surface.
A CT scan and stereophotography were performed. The 3D CT models were reconstructed by Materialise Mimics software, and the stereophotogrammetric facial surface was reconstructed by 3 dMD patient software. All 3D CT models were exported as Stereo Lithography file format, and the 3 dMD model was exported as Virtual Reality Modeling Language file format. Image registration and fusion were performed in Mimics software. Genetic algorithm was used for precise image fusion alignment with minimum error. The 3D CT models and the 3 dMD stereophotogrammetric facial surface were finally merged into a single file and displayed using Deep Exploration software. Errors between the CT soft tissue model and 3 dMD facial surface were also analyzed.
Virtual model based on CT-3 dMD image fusion clearly showed the photorealistic face and bone structures. Image registration errors in virtual face are mainly located in bilateral cheeks and eyeballs, and the errors are more than 1.5 mm. However, the image fusion of whole point cloud sets of CT and 3 dMD is acceptable with a minimum error that is less than 1 mm.
The ease of use and high reliability of CT-3 dMD image fusion allows the 3D virtual head to be an accurate, realistic, and widespread tool, and has a great benefit to virtual face model.
本研究的目的是通过三维（3D）CT模型与3dMD立体摄影面部表面的图像融合来证明构建颅面虚拟现实模型的可行性。
进行CT扫描和立体摄影。通过Materialise Mimics软件重建3D CT模型，通过3dMD患者软件重建立体摄影面部表面。所有3D CT模型均导出为立体光刻文件格式，3dMD模型导出为虚拟现实建模语言文件格式。在Mimics软件中进行图像配准和融合。使用遗传算法进行精确的图像融合对齐，误差最小。最终将3D CT模型和3dMD立体摄影面部表面合并为一个文件，并使用Deep Exploration软件显示。还分析了CT软组织模型与3dMD面部表面之间的误差。
基于CT-3dMD图像融合的虚拟模型清晰地显示了逼真的面部和骨骼结构。虚拟面部的图像配准误差主要位于双侧脸颊和眼球，误差超过1.5毫米。然而，CT和3dMD的全点云集的图像融合是可以接受的，最小误差小于1毫米。
CT-3dMD图像融合的易用性和高可靠性使3D虚拟头部成为一种准确、逼真且广泛应用的工具，对虚拟面部模型有很大益处。
