基于互信息的多分辨率三维脑图像配准方法

Method of Multi-Resolution 3D Image Registration by Mutual Information

在 3 D多模医学图像的配准方法中 ,最大互信息法精度高、鲁棒性强、使用范围广。本文将归一化互信息作为相似性测度 ,采用不同的采样范围和采样子集 ,使用 Powell多参数优化法和 Brent一维搜索算法对 3 D CT、MR和 PET脑图像进行了刚体配准。为了加快配准速度 ,使用了多分辨的金字塔方法。对 PET图像采用基于坐标的阈值选取方法对图像进行分割预处理 ,消除了大部分放射状背景伪影。

Maximization of mutual information is a powerful criterion for 3D medical image registration, allowing robust and fully accurate automated rigid registration of multi-modal images in a various applications. A method based on normalized mutual information for 3D image registration was presented on the images of CT, MR and PET. Powell's direction set method and Brent's one-dimensional optimization algorithm were used as optimization strategy. A multi-resolution approach is applied to speedup the matching process. For PET images, pre-procession of segmentation was performed to reduce the background artefact. According to the evaluation by the Vanderbilt University, Sub-voxel accuracy in multi-modality registration had been achieved with this algorithm.