CT脑血管图像三维重建中的层间目标对齐算法  

Object Aligning Algorithm for 3D CT Cerebrovascular Image Reconstruction

作　　者：雷印胜[1] 王明时[1] 秦然[2] 

机构地区：[1]天津大学精密仪器与光电子工程学院,天津300072 [2]山东青年管理干部学院信息工程系,济南250014

出　　处：《天津大学学报》2007年第9期1128-1133,共6页Journal of Tianjin University(Science and Technology)

摘　　要：在CT脑血管三维重建过程中,为了解决二维医学图像的层间目标对齐问题,提出了一种功效函数新算法.该算法根据图像的内部信息,一次性将断层图像移动到理想对齐位置的附近,然后在其目标周围寻找使相邻断层各匹配轮廓平均距离最小和功效函数最小的位置,即为最佳对齐位置.用几何中心代替质量重心,求出各轮廓的质量重心,降低了计算复杂度.A novel efficacy function algorithm was proposed in the process of the CT cerebrovascular 3-dimensional reconstruction to solve the paralleled serial slices object aligning problem of the 2-dimensional medical images. At first, based on the internal information of the images, the slices were moved to the place near the ideal position at a time. Then the optimal registration nearby which minimized the mean distance of every registration contour of adjacent slices and the efficacy function reflected the matching suitability at a given position. Every contour in a slice was designated a value that represented its size and density, then the overall centroid of the slice was found. Mass barycenter was replaced by the centroid, so that computation complexity at a large sCale was diminished.

关 键 词：CT脑血管医学图像 目标对齐 功效函数 三维重建 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]