MICCAI CLUST 2014 - bayesian real-time liver feature ultrasound tracking

We present the implementation of a Bayesian algorithm for tracking single features throughout ultrasound image sequences, with a focus on real-time applicability. After introducing the general concept of the algorithm, we suggest a sparse description of the target object to allow for rapid computation and semi-automatic target initialization. In 2D and 3D single feature tracking scenarios of the MICCAI challenge for liver ultrasound tracking (CLUST) 2014 we evaluate the algorithm and find mean tracking times of 1:25ms (2D) and 46:8ms (3D) per frame with mean tracking errors of 1:36mm (2D) and 2:79mm (3D).