A modified HAMMER algorithm for deformable registration of ultrasound images

A new technique for elastic registration of ultrasound images is proposed and evaluated. The algorithm is fast and robust, and has potential for intra-operative applications. It is a modified version of HAMMER technique, which was developed to elastically register magnetic resonance images of brain. A general concept of attribute vectors is employed to find correspondence across images, and new elements are suggested for the vectors to be able to distinguish between different features in ultrasound images. Registration results of simulated and actual deformation of ultrasound images of liver are presented.