基于光流和多层次B样条自由变形的医学图像鲁棒形变配准

为了跟踪肿瘤病灶对周围正常组织挤压变形，以及脑部手术软组织变形对手术的影响，提出了光流估计框架下的多层次B样条自由变形鲁棒形变配准方法.利用光流法约束下的多层次B样条自由变形，通过加密局部控制点以加强局部形变，能有效建模肿瘤信号及异常变形区域的局部较大形变.通过引入鲁棒估计，给予异常信号(病灶、灰度突变或形变不连续区域)以较小的权值，可减小图像灰度差异、局部不连续形变场所造成异常信号对形变配准优化的不利影响.采用了L BFGS方法，可减少内存开销，提高优化速度.实验表明：形变配准算法得到了较好的效果，可较好建模术前和术中脑部医学图像的形变，本配准算法可用于指导图像导向手术.

Robust Deformable Medical Image Registration Using Optical Flow and Multilevel Free Form Deformation

In order to track the brain tissue shift due to tumor resection during image guided surgery, this paper presented a framework for robust deformable registration based on the theory of optical flow, within which outlier detection is used to set up locally refined control points in multilevel free form deformation (FFD) model. Robust M estimator was also introduced to weaken the effect of outliers in global optimization for getting improved registration result. This proposed method has successfully been applied to a number of registration tasks to demonstrate its applicability and robustness to outliers.