基于弹性模型的CT与CARTO电解剖图像的配准算法

摘要

To propose an automatic and accurate CARTO electro-anatomic map and CT surface registration method.Coarse to fine strategy was used.Coarse registration was divided into two steps.First,using rigid transform and Iterative Closest Point (ICP) algorithm,EAM and CT surface were registered.Then,choosing affine transform,they were further aligned.Based on the coarse result as the initial point and B-spline free-form deformation as the transform model,two images were finely registered.Experiments show that our method alignes EAM and CT surface much more accurately and more steadily than Carto-Merge does ; meanwhile,no manual operation is needed during the whole coarse-to-fine procedure.%提出一种全自动、高精度的CARTO电解剖图与CT曲面配准算法.该算法采用了由粗到精的配准策略.粗配准部分分两步:首先采用刚体变换模型以及迭代最近点法,初次配准EAM与CT曲面；然后选择仿射变换模型,再次配准EAM与CT曲面.在粗配准的基础上,以基于B样条的自由形变模型进一步精确配准EAM与CT曲面.实验结果表明,相对临床常用的Carto-Merge配准软件,基于弹性模型的配准算法获得了远高于前者的EAM与CT曲面配准精度,配准效果稳定；同时算法完全自动,不需要任何手工介入.