A SPIRAL IN & OUT PULSE SEQUENCE DESIGN FOR RETROSPECTIVE CORRECTION SENSE

这个子项目是利用由NIH/NCRR资助的中心拨款提供的资源的许多研究子项目之一。子项目和调查员(PI)可能从另一个NIH来源获得了主要资金，因此可能会出现在其他CRISE条目中。列出的机构是针对中心的，而不一定是针对调查员的机构。简介运动伪影的校正仍然是MR中最基本的主题之一，特别是对于不合作的患者，如儿童和患有无法保持静止的疾病的患者，准确的运动确定和校正成为获得良好图像质量的必要条件。在这项研究中，我们提出了一种用于回溯运动校正的自旋回波螺旋输入输出序列，其目的是在平面内刚体运动(仅包括平移和旋转运动)的情况下消除与运动相关的伪影。为本研究设计的螺旋输入-输出序列可用于获取每次交错的低分辨率导航数据，而不会在扫描时间上造成额外的损失。改进的感测重建过程使用导航数据来寻找运动参数，并消除了k空间中欠采样的影响。材料与方法根据文献[1]中的算法，设计了一种自旋回波阿基米德螺旋输入输出脉冲序列。由于梯度系统的限制，螺旋弹道大多从摆率受限区域开始，在一定时间后切换到幅度受限区域，这取决于扫描参数。在本研究所使用的螺旋输入和输出轨迹的情况下，对于每个交错，使用螺旋输入轨迹来获得全采样的低分辨率图像，并且螺旋输出部分构成最终高分辨率图像的交错之一。该脉冲序列的一个优点是，螺旋部分利用180度脉冲之后的死区时间直到回波时间TE，并且这在T2加权的情况下不会对扫描时间带来损失。低分辨率导航仪数据的矩阵大小可以在操作员扫描之前交互地调整。这一序列已经在两名正常志愿者身上进行了测试，使用的是1.5T扫描仪(GE Signa LX，11.0)和高性能梯度系统(Gmax=50mt/m，SR=150mt/m/S)和8通道磁头阵列(MRI Devices)。志愿者被要求在扫描过程中每10秒将头部在头部线圈内移动大约10-20度，以模拟平面内刚体运动。所有人体研究都得到了我们机构审查委员会的批准。用于脉冲序列的其他参数如下：TR/TE=4000/56ms，层厚/间隙=5/0 mm，12层，FOV=24 cm，矩阵大小=256，交错=32，NEX=1，导航器矩阵大小=32，BW=125 kHz。扫描获得的数据被送入运动校正算法，该算法使用导航图像来完成联合配准，并获得旋转和平移量。在确定运动参数之后，通过对k空间轨迹进行反旋转并对k空间数据施加线性相位来相应地校正k空间轨迹、k空间数据和线圈灵敏度。这种运动校正在k空间中引入了一些间隙，并导致了图像域中的混叠。对于每个线圈和每个交错，使用广义Sense算法的修改版本来去除混叠和重建图像。结果给出了运动校正的结果。运动校正算法的应用，有效地消除了刚体运动产生的伪影。改进的SENSE算法通过填充k-空间轨迹的反旋转导致的k-空间中的空隙来改善最终的图像质量。从SENSE迭代中初始图像和最终图像之间的差异可以明显看出这一点。本研究中使用的螺旋轨迹允许更好地利用感觉重建，这是由于旋转校正后k空间中恒定的欠采样因子。在像EPI这样的其他轨迹的情况下，单个交错的旋转在k空间中留下了大的且任意间隔的间隙，这是无法用感官纠正的。有效缩减因子Reff被用作k空间欠采样的度量，其定义为两个螺旋臂之间的最大距离与原始k空间采样密度的比率。对于运动损坏的数据集，Reff为1.65，这是一个可以用SENSE纠正的合理值。显示了针对所有32个交错获得的导航器图像和运动校正图像。由于运动，对象暴露于每个交错的不同组合线圈敏感度。这导致导航器图像具有略微不同的强度变化，这可能会影响配准。感谢这项工作得到了NIH(1R01EB002771)、斯坦福高级磁共振技术中心(P41RR09784)、卢卡斯基金会和橡树基金会的部分支持。参考文献[1]Gever GH，MRM，42：412-415(1999)。[2]Atkinson D等，MRM，42：963-969(1999)[3]Pruessmann等，MRM，46：638-651(2001

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. Introduction Correction of motion artifacts still remains to be one of the most essential topics in MR. Especially in the case of uncooperative patients such as children and patients suffering from a medical condition that prevents them from staying stationary, accurate determination and correction of motion becomes a must for good image quality. In this study, we propose a spin-echo spiral in & out sequence for retrospective motion correction that is aimed to remove motion related artifacts in the case of in-planar rigid body motion, which includes only translational and rotational motion. The spiral in & out sequence designed for this study can be used to get low resolution navigator data for each interleave with no extra penalty in scan time. The modified SENSE reconstruction procedure uses that navigator data to find the motion parameters and eliminates the effects of undersampling in k-space. Materials and Methods A spin echo Archimedian spiral in & out pulse sequence is designed for this study according to the algorithm described in [1]. Due to the gradient system limitations, a spiral trajectory mostly starts off in slew rate limited region and switches to amplitude limited region after a certain time which is determined by the scan parameters. In the case of the spiral in & out trajectory used for this study, a spiral in trajectory is used to get a fully sampled low resolution image for each interleave, and the spiral out part constitutes one of the interleaves of the final high resolution image. One advantage of this pulse sequence is that the spiral in portion makes use of the dead time after the 180o degree pulse up to the echo time TE and this introduces no penalty for scan time in case of T2 weighting. The matrix size of the low resolution navigator data can be adjusted interactively before scan by the operator. This sequence has been tested on two normal volunteers using a 1.5T scanner (GE Signa LX, 11.0) with a high performance gradient system (Gmax = 50mT/m, SR = 150 mT/m/s) and an 8 channel head array (MRI Devices). The volunteers were asked to move their head inside the head coil by approximately 10-20 degrees for every 10 seconds during the scan to simulate in-planar rigid body motion. All human studies were approved by the review board of our institution. Other parameters used for the pulse sequence are as follows: TR/TE = 4000/56 ms, slice thickness/gap = 5/0 mm, 12 slices, FOV = 24 cm, matrix size = 256, interleaves = 32, NEX = 1, navigator matrix size = 32 and BW = 125 kHz. The data obtained from the scans were fed into a motion correction algorithm that uses the navigator images to accomplish co-registration and to obtain the amount of rotation and translation. After the determination of motion parameters, k-space trajectory, k-space data and the coil sensitivities are corrected accordingly by counter-rotating the k-space trajectories and applying a linear phase to k-space data. This motion correction introduces some gaps in k-space and causes aliasing in image domain. A modified version of the generalized SENSE algorithm that has a channel for each coil and for each interleave is used to remove aliasing and reconstruct the image. Results The results of motion correction are shown. The artifacts resulting from rigid body motion are significantly removed by the application of motion correction algorithm. The modified SENSE algorithm provides improvement in the final image quality by filling in the gaps in k-space resulting from the counter-rotation of k- space trajectories. This is apparent from the difference between the initial image and final image in the SENSE iteration. The spiral trajectory used in this study allows for better utilization of SENSE reconstruction due to the constant undersampling factor throughout k-space after rotation correction. In case of other trajectories like EPI, rotation of individual interleaf leaves large and arbitrarily spaced gaps in k-space which cannot be corrected by SENSE. An effective reduction factor, Reff, is used as a measure of the k-space undersampling and is defined as the ratio of the maximum distance between two spiral arms to the original k-space sampling density. For the motion corrupted data sets, Reff is 1.65, which is a reasonable value that can be corrected with SENSE. The navigator images obtained for all 32 interleaves and the motion corrected images are shown. Because of the motion, the subject is exposed to different combined coil sensitivity for each interleave. This results in navigator images having a slightly different intensity variation which might affect the registration. Acknowledgements This work was supported in part by the NIH (1R01EB002771), the Center of Advanced MR Technology at Stanford (P41RR09784), Lucas Foundation and Oak Foundation. References [1] Glover GH, MRM, 42:412-415 (1999). [2] Atkinson D, et al., MRM, 42:963-969 (1999) [3] Pruessmann et al, MRM, 46:638-651 (2001