基于ASM的CT图像序列标记点定位方法研究

CT图像序列中的标记点是CT图像配准和三维重建的基础,如何快速、准确地对标记点定位是一个值得研究的重要问题。论文引入活动轮廓模型(ASM)方法,并根据CT图像序列本身具有的相关性和实际特点,采用距离场插值方法获取迭代初始模型来解决ASM方法中初始模型获取问题,采用自适应尺度变换的检索方法来提高定位准确性和效率,并根据具体问题的需要采用灰度向量Mahalanobis距离最小作为匹配原则。试验验证,经过以上改进的ASM方法能够有效地解决CT图像序列中的标记点跟踪问题。

Using ASM to Locate the Mark Points in CT Image Series

The mark points in CT image series are base for image matching and 3_D image reconstruction.How to locate the mark points efficiently and exactly is a desirable question to research.This paper imports the ASM method,and implements some modifications on ASM method according to the correlations and characteristics of the CT image series.The modifications can be described as the following three points in general.Firstly,this paper adopts the interposition based on Distance Field theory to get the initial models.Secondly,using dynamic scales for searching the replace model.Thirdly,minimizing the Mahalanobis-Distance to select the replace points.The experiments prove that the modified ASM method can locate the mark points in CT image series effectively.