Correcting T2* effects in the myocardial perfusion arterial input function avoids overestimation of myocardial blood flow

Background Quantitative myocardial perfusion imaging requires correct measurement of arterial Gd concentration as the input function (AIF). An underestimation of AIF Gd concentration will lead to overestimation of myocardial blood flow (MBF). One source of this underestimation is the T2* effect due to high Gd concentration in the AIF measurement. Despite the short echo time (0.65 ms) used in the dual sequence approach specifically designed to avoid T2* loss, we have found significant loss remains at peak concentrations (5-10 mmol/l). To correct this signal loss in the AIF, a conventional dual sequence AIF acquisition [1] is modified to acquire two echoes per phase encode. The T2* effect can be corrected by modeling the exponential decay between two echoes. To validate the influences of T2* on the MBF estimates, pixel-wise MBF maps are computed using the Gadgetron [2] based inline perfusion flow mapping. For both rest and stress perfusion, the MBF values are compared with or without AIF T2* correction.