Ke Li1,2, Zhongliang Zu1,2, Junzhong Xu1,2, John C. Gore1,2, Heather M. Whitney1,3, Daniel F. Gochberg1,2
1Vanderbilt University Institute of Imaging Science, Nashville, TN, USA; 2Department of Radiology and Radiological Sciences, Vanderbilt University, Nashville, TN, USA; 3Department of Physics and Astronomy, Vanderbilt University, Nashville, TN, USA
Quantitative magnetization transfer (qMT) imaging maps properties of the tissues that are usually interpreted in terms of two pools of protons, corresponding to free and immobilized fractions. The selective-inversion-recovery fast-spin-echo (SIR-FSE) qMT technique developed recently includes an inversion time (ti) which is varied between 3.5 ms and 10 s, while the delay before the next sequence repetition (td) is held constant. qMT parameters are determined by fitting the resulting recovery to a bi-exponential function of ti using an approximate solution. In the current study, we employ a new protocol that varies both ti and td and fits the data with minimal approximations. Cramer-Rao lower bounds (CRLB) are calculated to select the variations in both ti and td that will maximize the precision-per-unit-time. Monte Carlo simulations support this approach by showing a large reduction in the resulting qMT parameter uncertainties. The optimization results are also confirmed by measurements on a series of BSA phantoms with different percent weight.