Similarity Measures for Non-Rigid Registration

Similarity measures for non-rigid multimodal registration are required to be local in order to enable correction of small image differences, and multimodal, to allow images to be acquired using different imaging techniques. Unfortunately all commonly used multimodal similarity measures are inherently global and cannot be directly used to estimate local image properties. We have derived a local similarity measure based on joint entropy, which can operate on extremely small image regions, e.g. individual voxels. The disadvantage of using such small image regions is higher sensitivity to noise and partial volume voxels, which reduce registration speed and accuracy. To cope with these problems we support the similarity measure with image segmentation. Several experiments based on synthetic images show that simultaneous application of registration and segmentation can improve registration accuracy and reduce the required number of registration steps.