Motion Tracking in Medical Images
Chuqing Cao, Chao Li, Ying Sun · 2015
This chapter introduces three types of motion tracking techniques: point-based tracking, silhouette-based tracking, and kernel-based tracking techniques, each with a most representative method, namely, Bayesian tracking method, deformable tracking model, and harmonic phase (HARP) algorithm. These three techniques are widely used for motion tracking in medical image now-a-days. A detailed case study of cardiac motion tracking in myocardial perfusion magnetic resonance imaging (MRI) is also included in the chapter. Bayesian tracking methods, such as Kalman filter and particle filter, are the commonly used methods for object tracking. Deformable models offer a unique and powerful approach to image analysis that combines geometry, physics, and approximation theory. The HARP method has become a popular alternative to object tracking for tagged medical images in recent years and it can be used to synthesize conventional tag lines, reconstruct displacement fields for small motions, and accurately track the object motion.