磁共振成像系统中的梯度线圈产生用于选层、频率编码和相位编码的梯度磁场.目前常用的梯度线圈是通过目标场法设计得到的.近些年来，由多个形状相同的线圈组成的矩阵式梯度线圈的梯度磁场均匀度和功率等指标也达到了较为满意的效果.本文首先提出了一种基于粒子群与遗传算法的、适用于开放式永磁型磁共振成像系统的矩阵式梯度线圈设计方法.然后对三个方向上的矩阵梯度线圈的电流分布进行了设计，每个方向上的矩阵式梯度线圈系统由224个大小相同的圆形线圈组成.最后利用有限元仿真软件对设计方案进行仿真计算，得到x、y方向上的平均非均匀度为0.851%，z方向上的平均非均匀度为1.013%，验证了本文提出的方法的有效性.基于该方法可以有效快速地对开放式磁共振成像系统的矩阵梯度线圈进行设计.
王强
,
魏树峰
,
王铮
,
杨文晖
. 基于粒子群与遗传算法的矩阵式梯度线圈优化设计[J]. 波谱学杂志, 2019
, 36(4)
: 463
-471
.
DOI: 10.11938/cjmr20192730
Gradient coils generate gradient magnetic fields that are essential in the magnetic response imaging. The gradient coils are usually designed by the target field method. In this work, we attempted to design matrix gradient coils with particle swarm optimization and the genetic algorithm. Finite element simulation calculation demonstrated that the average non-uniformity in the x and y directions was 0.851%, and the average non-uniformity in the z direction was 1.013%.
[1] TURNER R. Gradient coil design:A review of methods[J]. Magn Reson Imaging, 1993, 11(7):903-920.
[2] TURNER R. A target field approach to optimal coil design[J]. J Phys D Appl Phys, 1986, 19(8):L147-L151.
[3] MANSFIELD P, CHAPMAN B L, BOWTELL R, et al. Active acoustic screening:reduction of noise in gradient coils by Lorentz force balancing[J]. Magn Reson Med, 1995, 33(2):276-281.
[4] YODA K. Analytical design method of self-shielded planar coils[J]. J Appl Phys, 1990, 67(9):4349-4353.
[5] CHRISTOPH J, NIXON T W, SCOTT M I, et al. Magnetic field modeling with a set of individual localized coils[J]. J Magn Reson, 2010, 204(2):281-289.
[6] JIA F, SCHULTZ G, TESTUD F, et al. Performance evaluation of matrix gradient coils[J]. MAGMA, 2016, 29(1):59-73.
[7] JIA F, LITTIN S, LAYTON K J, et al. Design of a shielded coil element of a matrix gradient coil[J]. J Magn Reson, 2017, 281:217-228.
[8] KROBOTH S, LAYTON K J, JIA F, et al. Optimization of coil element configurations for a matrix gradient coil[J]. IEEE T Med Imaging, 2018, 37(1):284-292
[9] SHI X H, WAN L M, LEE H P, et al. An improved genetic algorithm with variable population-size and a PSO-GA based hybrid evolutionary algorithm[C]//International Conference on Machine Learning & Cybernetics, 2003. doi:10.1109/ICMLC.2003.1259777.
