中文摘要：

目的：研究利用初步建立的三维模型确定二维断层图像感兴趣结构像素坐标的方法。方法：通过Photoshop图像处理软件绘制断面图像．使用可视化工具包VTK的移动立方体表面重建算法，在VC＋＋6．0的编译环境下对其进行三维重建以及立体显示。用自行开发的坐标转换处理程序对三维模型上提取的坐标值进行处理，计算断层图像相应结构的像素坐标。结果：建立了一个表面带有S型凹槽的三维模型，通过计算三维模型上凹槽结构的一系列坐标，得到二维断层图像上相应结构的像素坐标点。结论：本研究以VTK重建的三维模型为基础，提出了一种利用已建成的三维模型来指导二维断层图像结构定位的方法．为人体复杂结构的分割与修正以及某些在二维断层图像上无法识别的结构的定位提供了一种新的手段。

英文摘要：

Objective： To study the methods of using the initial 3D model to orientate the interesting flame＇s pixel coordinates on the sectional images. Methods： To draw a serial of sectional images by Photoshop software. To construct these 2D contours into a 3D visual model by metals of the surface reconstruction algoritlan programmed in VTK under VC＋＋ IDE. We reconstructed and viewed the 3D model and calculated the corresponding pixel coordinates on the sectional images according to the coordinates on the 3D model by using the processor software which was developed by our research group. Results： A 3D model with a S-shaped fillister on its surface was reconstructed. The corresponding pixel coordinates of this fillister on the 2D sectional images were obtained by calculating its coordinates on the 3D model. Conclusions： This research proposed a new method to allow fine segmentation of the 2D sectional image by using established 3D model in VTK. This method would be helpful to refine the coarse segmentation of the complex anatomic structure of the human body, and especially realize the segmentation of some unrecognized structure on 2D sectional images.