ICP Registration Technology based on the Coordinate System Direction Fit

In computer assisted surgery, one of the most important problems is to align the preoperative model with intraoperative data. The Iterative Closest Point (ICP) algorithm is undoubtedly the most popular algorithm for solving this kind of problem, which needs a good initialization. In this paper, an ICP algorithm based on the coordinate system direction fit is present to improve ICP algorithm's shortcoming with easily falling into local optimum for intraoperative registration. Firstly, ICP registration algorithm and its problems are introduced. Secondly, based on the consistency between CT scans pose and surgical pose, a three-plane coordinate system on patients is established to reduce the rotation dislocation between CT image space and patient space. Then, an ICP algorithm experiment and its precision analysis are conducted. Finally, the error analysis is also given. The experimental results show that one of the benefits of the ICP algorithm based on the coordinate system direction fit is that it is easily convergent on the registration by a three-plane coordinate system on patients, which increases the registration accuracy and reduces the computational processing time