为避免成像物体在核磁共振成像系统(MRI)实际操作中的旋转难题,基于微分进化思想提出了一种3维磁共振电阻抗成像算法(MREIT),并采用单方向磁感应强度"测量值"在3维真实头模型上进行了仿真验证研究。结果表明:该算法可以对病变的真实头模型进行阻抗图像重构,且计算值和理论值之间的相对误差〈5%,具有较高的成像精度和空间分辨率;实时调整参数因子以适应算法,可以缩短重构时间,加快算法收敛性。因此改进后的算法具有很好收敛性,且成像质量较高,在临床应用上具有一定的可行性。
In order to overcome the difficulty of rotating imaging objects in the operation of magnetic resonance imaging(MRI) system, we proposed an algorithm for three-dimensional magnetic resonance electrical impedance tomography(MREIT) by the differential evolution theory. This algorithm was verified by using a three-dimensional realistic-geometry head model and single-direction magnetic-flux-density measured data. The results show that the proposed algorithm can reconstruct the impedance image of diseased realistic-geometry head model. The error between the calculated value and theoretical value is less than 5%; this indicates high imaging precision and high spatial resolution of the proposed algorithm. By improving factors in real time that fits the algorithm better, it is able to shorten the reconstruction time and accelerate the algorithm convergence. It is concluded that the improved algorithm has good convergence and high imaging quality, and it has potential in clinical applications.