RR-UNFOLD: Respiratory Reordered UNFOLD for First Pass Myocardial Perfusion Imaging

Early diagnosis and localisation of myocardial perfusion defects is an important step in the treatment of coronary artery disease. The assessment of myocardial perfusion requires accurate quantification of the transmural extent of possible defects, which involves a complete multi-slice coverage of the ventricle at a given phase of the cardiac cycle. To this end, a number of rapid imaging sequences have been proposed in recent years [1] to minimise the data acquisition window so as to avoid misregistration due to cardiac as well as respiratory motion. UNFOLD [2] is an image acquisition and reconstruction method which attempts to encode spatial information into redundant regions of k-t space. The sub sampling of k-space results in aliasing in the spatial domain and the success of the method is dependant on dynamic regions being uniquely represented in the spatial domain. Since myocardial perfusion imaging typically involves 50 cardiac cycles, respiratory induced cardiac deformation imposes a major limitation to the application of the UNFOLD method [3]. This paper presents a novel approach to the acquisition and reconstruction of MR myocardial perfusion images based on UNFOLD but with prospective respiratory phase encode reordering. It provides an adaptive real-time binning method that minimises the effect of respiration whilst maintaining the temporal tissue characteristics during the contrast up-take.