ABSTRACT

MRI tagging has been used in various medical research investigations. It has been used to develop models of normal and abnormal myocardial motion (Clark et  al. 1991, McVeigh and Zerhouni 1991, Moore et  al. 1992, Young and Axel 1992, McVeigh 1996) and to better understand myocardial motion abnormalities associated with ischemic heart disease (Lima et  al. 1995) and heart electrophysiology (McVeigh et  al. 1998). Most approaches implemented for MRI tag analysis used image processing techniques, for example, deformable models (Young and Axel 1992) and matched lters (Guttman et  al. 1994), to detect and track the tagging pattern deformation during the cardiac cycle and generate a detailed motion map using nite element models (FEMs) (Young and Axel 1992), B-splines (Radeva et al. 1996), or statistical models (Denney and Prince 1995). Nevertheless, the long processing time of these techniques and lack of fast quantitative analysis and visualization techniques are preventing MRI tagging from being widely adopted in the clinical setting. The harmonic phase (HARP) technique, introduced in this chapter, stands different from the techniques mentioned earlier as it is based on analyzing the frequency domain (k-space) of the tagged images instead of analyzing the images themselves, which results in more automated analysis and faster processing. HARP allows rapid analysis and visualization of myocardial strain with processing time an order of magnitude faster than conventional analysis techniques.