Rapid and accurate dictionary-based T 2 mapping from multi-echo turbo spin echo data at 7 Tesla.

Background Using lower refocusing flip angles in multi-echo turbo spin echo (ME-TSE) sequences at ultra-high magnetic field leads to non-monoexponential signal decay and overestimation of T-2 values due to stimulated and secondary echoes. Purpose To investigate the feasibility of a fast and accurate reconstruction of quantitative T-2 values using an ME-TSE sequence with reduced refocusing flip angles at 7 Tesla, a dictionary-based reconstruction method was developed and is presented in this work. Study Type Prospective. Subjects Phantom measurements with relaxation phantom, four healthy volunteers. Field Strength/Sequence 7 Tesla MRI, multi-echo turbo spin echo (ME-TSE), spin echo (SE), and B-1 mapping. Assessment Based on Bloch simulations and the extended phase graph model, signal decay curves were calculated to account for nonrectangular slice profile, B-1 inhomogeneity, and reduced refocusing flip angles and stored in a dictionary. Data obtained with an ME-TSE sequence at 7 Tesla were matched to this dictionary to obtain T-2 values. To compare the proposed method to reference T-2 values, a spin echo sequence with different echo times was used. Statistical Tests Welch's t-test was used to compare T-2 values in phantom measurements. Results T-2 values obtained with the proposed ME-TSE method coincided with the T-2 values from the spin echo experiment in phantom measurements (P = 0.89 for 120 degrees flip angle, P = 0.75 for 180 degrees flip angle). Only for very low B-1 transmit fields, a slight overestimation of T-2 values was observed. In vivo measurements showed lower T-2 values in gray matter (55 +/- 2 millisecond) and white matter (39 +/- 5 millisecond) compared with literature values of 3 Tesla data. Data Conclusions The proposed dictionary-based ME-TSE approach provided accurate T-2 values in short measurement time at 7 Tesla with low specific absorption rate burden due to the reduction of refocusing flip angles. Therefore, it can provide new opportunities in clinical high-field MRI to further improve radiographic diagnosis by using quantitative imaging.