开放式自屏蔽全身成像高场超导MRI磁体优化设计

本文发展了开放式自屏蔽全身成像高场超导磁共振成像（MRI）磁体的优化设计方法，使设计出来的磁体仅有4 对超导线圈. 这种开放结构的超导MRI磁体优化设计方法集成了线性规划算法和遗传算法. 通过迭代线性规划算法可以在考虑成像区域（DSV）磁感应强度约束、磁场不均匀度约束、5 Gs线范围约束、线圈区域最大磁场值约束和最大环向应力约束的条件下，获得用线量最少的线圈初始形状和位置，同时可以得到每个线圈的层数和每层匝数；通过遗传算法可以提高DSV区域的磁场不均匀度，以达到高质量成像的要求. 这种集成的优化设计方法既可以灵活有效的设计开放式MRI磁体，也可以设计传统的圆柱形MRI磁体，本文通过一个1.2 T的开放式MRI磁体的设计清楚的展示了这种优化方法. 关键词： 线性规划算法 遗传算法 自屏蔽 开放式超导MRI磁体

Optimization in design of actively shielded whole-body open high-field superconducting MRI magnet

We propose an optimization design method for actively shielded whole-body open high-field superconducting MRI magnet which accordingly has been simplified to contain only four pairs of superconducting coils. The new design method of open structure superconducting MRI magnet integrates the linear programming algorithm and the genetic algorithm optimization. Through several times of linear programming, and by taking into account the magnetic field, inhomogeneity in DSV, the scope of 5-Gauss fringe field, maximum hoop stress and maximum magnetic field, we can, with the least consumption of lines, get the coils’ initial position and shape, the number of layers of each coil and the number of turns of each layer. And the genetic algorithm was then employed to improve the magnetic field inhomogeneity in DSV to meet the requirements of high-quality imaging. This integrated optimization design method is flexible and effective for designing both open MRI magnet and traditional cylindrical MRI magnet. This paper also illustrates the method for a 1.2 T open MRI magnet optimization design.
Keywords： linear programming algorithm genetic algorithm actively shielded open superconducting MRI magnet