Electrical and Computer Engineering, University of Minnesota, Minneapolis, Minnesota, USA.
Center for Magnetic Resonance Research, University of Minnesota, Minneapolis, Minnesota, USA.
Magn Reson Med. 2017 Aug;78(2):462-471. doi: 10.1002/mrm.26770. Epub 2017 Jun 5.
To develop a saturation recovery myocardial T mapping method for the simultaneous multislice acquisition of three slices.
Saturation pulse-prepared heart rate independent inversion recovery (SAPPHIRE) T mapping was implemented with simultaneous multislice imaging using FLASH readouts for faster coverage of the myocardium. Controlled aliasing in parallel imaging (CAIPI) was used to achieve minimal noise amplification in three slices. Multiband reconstruction was performed using three linear reconstruction methods: Slice- and in-plane GRAPPA, CG-SENSE, and Tikhonov-regularized CG-SENSE. Accuracy, spatial variability, and interslice leakage were compared with single-band T mapping in a phantom and in six healthy subjects.
Multiband phantom T times showed good agreement with single-band T mapping for all three reconstruction methods (normalized root mean square error <1.0%). The increase in spatial variability compared with single-band imaging was lowest for GRAPPA (1.29-fold), with higher penalties for Tikhonov-regularized CG-SENSE (1.47-fold) and CG-SENSE (1.52-fold). In vivo multiband T times showed no significant difference compared with single-band (T time ± intersegmental variability: single-band, 1580 ± 119 ms; GRAPPA, 1572 ± 145 ms; CG-SENSE, 1579 ± 159 ms; Tikhonov, 1586 ± 150 ms [analysis of variance; P = 0.86]). Interslice leakage was smallest for GRAPPA (5.4%) and higher for CG-SENSE (6.2%) and Tikhonov-regularized CG-SENSE (7.9%).
Multiband accelerated myocardial T mapping demonstrated the potential for single-breath-hold T quantification in 16 American Heart Association segments over three slices. A 1.2- to 1.4-fold higher in vivo spatial variability was observed, where GRAPPA-based reconstruction showed the highest homogeneity and the least interslice leakage. Magn Reson Med 78:462-471, 2017. © 2017 International Society for Magnetic Resonance in Medicine.
开发一种用于同时采集 3 个层面的饱和恢复心肌 T 映射方法。
使用 FLASH 读出实现饱和脉冲准备的心率独立反转恢复(SAPPHIRE)T 映射,用于更快地覆盖心肌。使用并行成像的受控消隐(CAIPI)实现 3 个层面的最小噪声放大。使用 3 种线性重建方法进行多频带重建:层间和平面 GRAPPA、CG-SENSE 和 Tikhonov 正则化 CG-SENSE。在体模和 6 名健康志愿者中,与单频带 T 映射比较准确性、空间变异性和层间渗漏。
多频带体模 T 时间与所有 3 种重建方法的单频带 T 映射具有良好的一致性(归一化均方根误差<1.0%)。与单频带成像相比,GRAPPA 的空间变异性增加最低(1.29 倍),而 Tikhonov 正则化 CG-SENSE(1.47 倍)和 CG-SENSE(1.52 倍)的惩罚更高。与单频带(T 时间±节段间变异性:单频带,1580±119 ms;GRAPPA,1572±145 ms;CG-SENSE,1579±159 ms;Tikhonov,1586±150 ms[方差分析;P=0.86])相比,多频带 T 时间在体内无显著差异。层间渗漏最小的是 GRAPPA(5.4%),而 CG-SENSE(6.2%)和 Tikhonov 正则化 CG-SENSE(7.9%)更高。
多频带加速心肌 T 映射显示了在 3 个层面上进行 16 个美国心脏协会节段单次屏气 T 定量的潜力。观察到体内空间变异性增加 1.2-1.4 倍,其中基于 GRAPPA 的重建显示出最高的均匀性和最小的层间渗漏。磁共振医学 78:462-471,2017。© 2017 国际磁共振学会。