Feasibility of T 2 * Estimation with Chemical Shift-Based Water-Fat Separated Cardiac Imaging

K. K. Vigen,H. Yu,C. J. Francois,A. Shimakawa,S. B. Reeder
2009-01-01
Abstract:Introduction T2* mapping has been investigated in cardiac MRI for imaging the myocardial BOLD effect [1] and in iron overload assessment [2]. Advanced T2* estimation techniques in the context of chemical shift-based water-fat decomposition have recently been developed for iron detection in organs such as the liver, while simultaneously estimating fat content [3]. Separately, chemical shiftbased water-fat decomposition techniques are being developed for cardiac MRI [4-6]. Myocardial T2* measurements that account for the effects of fat on the T2* decay curve have been described [7]. In this work, we investigate the feasibility of chemical shift-based water-fat decomposition using an advanced IDEAL T2* estimation method with fat spectral peak correction [3]. Methods A cardiac-gated, breath-held multi-echo gradient echo sequence with 10 positive (fly-back) readout echoes was implemented at 1.5T (Signa HDx, GE Healthcare, Waukesha, WI) using an 8-channel phased array cardiac coil. Acquisition parameters included TR/TE1/ΔTE/BW = 27.3ms/1.6ms/1.8ms/±100kHz; FOV/Matrix/Slice Thickness = 35cm/192x192/8mm; 16 views per segment; and one segment every 2 cardiac cycles, requiring 25 cardiac cycles per slice. Imaging was performed in five normal volunteers; informed consent was obtained according to a protocol approved by the local Institution Review Board. For each acquisition, three short-axis LV slices were acquired, one each at basal, mid-ventricular, and apical locations. Waterand fat-separated images were generated with an investigational version of 10-pt IDEAL algorithm [8-9], and an advanced method for simultaneous multi-spectral water-fat decomposition and T2* estimation [3] was used to generate T2*-corrected water and fat images and R2* maps on-line. Results Figure 1 shows mid-ventricular (Fig. 1a-d) and apical (Fig. 1e-h) slice locations in one volunteer. Conventional IDEAL images (without T2* correction), are effectively a linear combination of all 10 echo source images and show some T2* weighting (Fig. 1a,e) and superior SNR. The T2* contrast was corrected with the advanced algorithm [3] (Fig. 1b,f), with estimated R2* maps (Fig. 1c,g). Fat-only images (Fig. 1d,h) helped demonstrate the excellent water-fat separation.
What problem does this paper attempt to address?