Fast Imaging of Cardiac Strain Using Partial k-Space HARP in Mice

Characterization of myocardial deformation in genetically manipulated mouse models with MR tagging provides new opportunities for elucidating the molecular mechanisms of cardiac function. However, tagging on murine hearts presents additional challenge because of the requirement for high tagging and imaging resolution. The high heart rate of mice renders EPI or segmented k-space sampling impractical. As a result, typical acquisition time for mice is much longer than that for humans. One direct consequence of long acquisition time is that more hemodynamic variations might occur during imaging acquisition that could potentially compromise image quality and complicate data interpretation. Recently, an ultra-fast acquisition method was developed based on Harmonic Phase (HARP) technique [1]. Tagged image is composed of multiple spectral peaks in the Fourier domain (Figure 1, left). The phase image of the inverse Fourier transformed off-center peaks directly encodes the motion of the underlying tissue [2]. Based on this property, ultra-fast HARP method was proposed to sample only the off-center spectral peaks. However, direct reconstruction of HARP image from the kspace data of a single off-center peak suffers from inaccuracies in phase correction. Phase correction is further complicated by the use of surface coil and high imaging/tagging gradients in small animal imaging to maximize signal-to-noise ratio and achieve high spatial and temporal resolution. Here we propose a partial k-space sampling approach for imaging mouse heart. Such an approach samples all the spectral peaks in the low-frequency part of the k-space. HARP analysis is based on the reconstructed magnitude images so that phase correction is not necessary. Validation was performed in the current study on four mice. The results exhibited excellent correlations with the full k-space sampling method.