非笛卡尔并行磁共振成像重建技术研究新进展

发布时间：2018-03-26 15:18

本文选题：磁共振成像技术　切入点：非笛卡尔采样　出处：《仪器仪表学报》2017年08期

【摘要】：相对于传统的k空间笛卡尔采样,非笛卡尔采样能够使得k空间具有更高的覆盖效率,同时可以更有效地利用梯度系统性能,减少d B/dt值,避免引起人体不良的生理反应。k空间非笛卡尔采样和并行成像技术结合能够进一步提高成像速度,但是也使得图像域中的伪影模式更加复杂,因此非笛卡尔并行磁共振成像重建具有更高的技术难度。综述了目前几种典型的非笛卡尔并行成像重建技术,具体讨论了每种方法的技术细节和优缺点,包括敏感度编码(SENSE)、共轭梯度敏感度编码(CGSENSE)、非笛卡尔自标定并行采集方法(non-Cartesian GRAPPA)、基于数据一致性的迭代方法(SPIRi T)和近年来发展迅速的压缩感知技术。SENSE和CG-SENSE理论上可以获得最优的重建结果,但受制于线圈敏感度分布的准确测量;Non-Cartesian GRAPPA无需测量线圈敏感度,但只能对特定的非笛卡尔采样模式进行近似计算;SPIRi T结合了SENSE和GRAPPA的优点,通过迭代优化方法可以获得较满意的结果;压缩感知技术利用图像的稀疏变换特性,配合现有的迭代优化并行成像方法可以进一步提升重建图像质量,将继续成为未来研究的热点。
[Abstract]:Compared with the traditional k-space Cartesian sampling, the non-Cartesian sampling can make the k-space have higher coverage efficiency, at the same time, it can make more effective use of the gradient system performance and reduce the d B/dt value. The combination of non-Cartesian sampling and parallel imaging technology can further improve the imaging speed, but it also makes the artifact mode in the image domain more complicated. Therefore, non-Cartesian parallel magnetic resonance imaging reconstruction is more difficult. Several typical non-Cartesian parallel imaging reconstruction techniques are reviewed, and the technical details, advantages and disadvantages of each method are discussed in detail. These include sensitivity coding, conjugate gradient sensitivity coding, CGSENSEN, non-Cartesian GRAPPAA, non-Cartesian GRAPPAA, iterative methods based on data consistency (SPIRI T) and the rapidly developed compression sensing techniques .SENSE and CG-SENSE in theory. To get the best result of reconstruction, However, accurate measurement of coil sensitivity distribution by non-Cartesian GRAPPA does not require the measurement of coil sensitivity, but can only be approximately calculated for a particular non-Cartesian sampling mode. SPIRI T combines the advantages of SENSE and GRAPPA. Through iterative optimization method, satisfactory results can be obtained, and compression sensing technology can further improve the quality of reconstructed images by utilizing the sparse transformation characteristics of images and combining with existing iterative optimization parallel imaging methods. Will continue to be the focus of future research.
【作者单位】：中国科学院苏州生物医学工程技术研究所医学影像室;
【基金】：中科院科研装备研制项目(YZ201445) 国家自然科学基金(11505281)项目资助
【分类号】：R445.2;TP391.41

【相似文献】