论文部分内容阅读
目的:通过增强现实技术,将下颌骨的虚拟三维数字模型及术前设计的预截骨平面同时显示在快速成型实体模型上,实现虚拟图像与实体的重叠配准。方法:选取20例下颌骨肥大患者,行三维CT扫描,重建下颌骨三维数字模型。利用上海九院整形设计系统完成术前设计,生成截骨平面,并与下颌骨合并为STL文件。取患者下牙石膏牙模,制作包含下牙咬合板在内的标志物支架,固定于石膏牙模,两者一并扫描。然后将扫描数据和下颌骨三维数字模型数据导入同一三维处理平台,选择右第一磨牙的远中舌尖、第二磨牙的近中舌尖、左第一磨牙的近中舌尖,进行标志物支架和下颌骨的拟合,生成虚拟影像。此时的三维虚拟数字化影像包括标志物、下颌骨及预截骨平面。然后将标志物支架的下牙咬合板固定于下颌骨快速成型模型上,采用增强现实视频检测方法,用视频捕捉器识别到标志物后,将虚拟影像与下颌骨快速成型实体模型进行配准。结果:该技术实现了三维虚拟数字化影像与下颌骨快速成型模型的虚实融合叠加,使下颌骨和术前设计的预截骨平面实时显示在下颌骨实体上。结论:本研究建立的配准方法具有良好的重复性,有望成为下颌角截骨术可视化手术有效的配准途径,为未来增强现实手术应用研究奠定了基础。
OBJECTIVE: To display the virtual three-dimensional digital model of the mandible and the pre-osteotomy plane preoperatively displayed on the rapid prototyping solid model through augmented reality technology to realize the overlay registration of virtual images and entities. Methods: Twenty patients with mandibular hypertrophy were selected and underwent three-dimensional CT scan. The three-dimensional digital model of the mandible was reconstructed. The use of Shanghai nine hospital plastic surgery design system to complete the preoperative design, generate osteotomy plane, and combined with the mandible STL file. Patient under the plaster dental model, including the lower occlusal plate containing the marker scaffold, fixed in the gypsum dental model, both scanned together. Then, the scan data and the three-dimensional digital model data of the mandible were introduced into the same three-dimensional processing platform. The distal tip of the right first molar, the near-middle tip of the second molar and the near-mid-tongue of the left first molar were selected for the marker scaffold and mandible Fitting the bone to create a virtual image. The three-dimensional virtual digital images at this time include markers, mandibular and pre-osteotomy planes. Then, the lower occlusal plate of the marker scaffold was fixed on the mandibular rapid prototyping model. After the augmented reality video detection method was used to identify the markers with the video capture, the virtual image and mandibular rapid prototyping solid model were registered. Results: The technique realizes the superposition of the virtual 3D model of the virtual digitalized image and the mandibular rapid prototyping model so that the pre-osteotomy plane of the mandible and preoperative design can be displayed on the mandible entity in real time. Conclusion: The registration method established in this study has good repeatability and is expected to be an effective registration method for visual surgery of mandibular angle osteotomy, which lays the foundation for the application of augmented reality surgery in the future.