Computer Science · Computer Vision and Pattern Recognition
Using K-means Clustering and MI for Non-rigid Registration of MRI and CT ∗
Yixun Liu, Nikos Chrisochoides
See on the mapAbstract
Abstract. Mutual information (MI) based registration methods are susceptible to the variation of the intensity of the image. We present a multi-modality MRI-CT non-rigid registration method by combining K-means clustering technique with mutual information. This method makes use of K-means clustering to determine variant bin sizes in CT image. The resulting clustered (labeled) CT image is non-rigidly registered with MRI by modeling the underlying movement as Free-Form Deformation (FFD). We compare this Cluster-to-Image registration method with Image-to-Image and Cluster-to-Cluster methods. The preliminary experiment shows this method can increase the accuracy of non-rigid registration. 1