Time Efficient DT-MRI Acquisition Parameters for Robust Estimation of Fiber Tracts

We investigate optimal diffusion tensor imaging (DTI) acquisition parameters including the number of diffusion encoding directions (K), the number of excitations at each direction (Ne) and diffusion weighting (b-value). A time efficiency measure is proposed for parameter optimization to achieve the best compromise between the accuracy of estimated fiber tracts and scan time in spin echo single shot echo-planar imaging (EPI). By simulation with four synthetic datasets, we conclude that six encoding directions and one excitation give the highest time efficiency when single image SNR(0) measured at b-value = 191.14 mm(-2) s is larger than 32. The optimal number of excitations increases to two when SNR0 is smaller than 32. Further increase of K and Ne would decrease time efficiency. And the optimal b-value decreases as SNR0 increases. Finally, we validate the findings by comparing the time efficiency of two sets of acquisition parameters using real dataset, and marked lift is observed when the optimal set is applied.