Department of Radiology, Dartmouth-Hitchcock Medical Center, Lebanon, NH 03756, USA.
Phys Med Biol. 2012 Nov 21;57(22):7275-87. doi: 10.1088/0031-9155/57/22/7275. Epub 2012 Oct 18.
Many pathologies alter the mechanical properties of tissue. Magnetic resonance elastography (MRE) has been developed to noninvasively characterize these quantities in vivo. Typically, small vibrations are induced in the tissue of interest with an external mechanical actuator. The resulting displacements are measured with phase contrast sequences and are then used to estimate the underlying mechanical property distribution. Several MRE studies have quantified brain tissue properties. However, the cranium and meninges, especially the dura, are very effective at damping externally applied vibrations from penetrating deeply into the brain. Here, we report a method, termed 'intrinsic activation', that eliminates the requirement for external vibrations by measuring the motion generated by natural blood vessel pulsation. A retrospectively gated phase contrast MR angiography sequence was used to record the tissue velocity at eight phases of the cardiac cycle. The velocities were numerically integrated via the Fourier transform to produce the harmonic displacements at each position within the brain. The displacements were then reconstructed into images of the shear modulus based on both linear elastic and poroelastic models. The mechanical properties produced fall within the range of brain tissue estimates reported in the literature and, equally important, the technique yielded highly reproducible results. The mean shear modulus was 8.1 kPa for linear elastic reconstructions and 2.4 kPa for poroelastic reconstructions where fluid pressure carries a portion of the stress. Gross structures of the brain were visualized, particularly in the poroelastic reconstructions. Intra-subject variability was significantly less than the inter-subject variability in a study of six asymptomatic individuals. Further, larger changes in mechanical properties were observed in individuals when examined over time than when the MRE procedures were repeated on the same day. Cardiac pulsation, termed intrinsic activation, produces sufficient motion to allow mechanical properties to be recovered. The poroelastic model is more consistent with the measured data from brain at low frequencies than the linear elastic model. Intrinsic activation allows MRE to be performed without a device shaking the head so the patient notices no differences between it and the other sequences in an MR examination.
许多病理学改变了组织的力学性质。磁共振弹性成像（MRE）已经被开发出来，用于无创地在体内对这些量进行特征描述。通常，使用外部机械致动器在感兴趣的组织中产生小的振动。使用相位对比序列测量产生的位移，然后用于估计潜在的机械特性分布。几项 MRE 研究已经量化了脑组织特性。然而，颅骨和脑膜，特别是硬脑膜，非常有效地阻止外部施加的振动穿透大脑深处。在这里，我们报告了一种方法，称为“固有激活”，它通过测量自然血管搏动产生的运动来消除对外界振动的要求。使用回顾性门控相位对比磁共振血管造影序列记录了心脏周期八个相位的组织速度。通过傅里叶变换对速度进行数值积分，以产生大脑内每个位置的谐波位移。然后将位移重建为基于线性弹性和多孔弹性模型的剪切模量图像。所产生的力学性质在文献中报道的脑组织估计范围内，同样重要的是，该技术产生了高度可重复的结果。线性弹性重建的平均剪切模量为 8.1 kPa，多孔弹性重建的平均剪切模量为 2.4 kPa，其中流体压力承担部分应力。大脑的大体结构得到了可视化，特别是在多孔弹性重建中。在对六名无症状个体的研究中，个体内的平均剪切模量的变异性明显小于个体间的变异性。此外，当在个体中随时间观察时，观察到机械特性的变化比在同一天重复 MRE 程序时更大。称为固有激活的心脏搏动产生足够的运动，允许恢复机械特性。多孔弹性模型比线性弹性模型更符合大脑在低频下的测量数据。固有激活允许在不摇动头部的设备的情况下进行 MRE，因此患者在磁共振检查中不会注意到它与其他序列之间的差异。
