Gao Nuo, Zhu S A, He Bin
College of Electrical Engineering, Zhejiang University, Hangzhou 310027, People's Republic of China.
Phys Med Biol. 2005 Jun 7;50(11):2675-87. doi: 10.1088/0031-9155/50/11/016. Epub 2005 May 18.
We have developed a new algorithm for magnetic resonance electrical impedance tomography (MREIT), which uses only one component of the magnetic flux density to reconstruct the electrical conductivity distribution within the body. The radial basis function (RBF) network and simplex method are used in the present approach to estimate the conductivity distribution by minimizing the errors between the 'measured' and model-predicted magnetic flux densities. Computer simulations were conducted in a realistic-geometry head model to test the feasibility of the proposed approach. Single-variable and three-variable simulations were performed to estimate the brain-skull conductivity ratio and the conductivity values of the brain, skull and scalp layers. When SNR = 15 for magnetic flux density measurements with the target skull-to-brain conductivity ratio being 1/15, the relative error (RE) between the target and estimated conductivity was 0.0737 +/- 0.0746 in the single-variable simulations. In the three-variable simulations, the RE was 0.1676 +/- 0.0317. Effects of electrode position uncertainty were also assessed by computer simulations. The present promising results suggest the feasibility of estimating important conductivity values within the head from noninvasive magnetic flux density measurements.
我们开发了一种用于磁共振电阻抗断层成像（MREIT）的新算法，该算法仅使用磁通密度的一个分量来重建体内的电导率分布。在本方法中，使用径向基函数（RBF）网络和单纯形法，通过最小化“测量的”和模型预测的磁通密度之间的误差来估计电导率分布。在逼真几何形状的头部模型中进行了计算机模拟，以测试所提出方法的可行性。进行了单变量和三变量模拟，以估计脑-颅骨电导率比以及脑、颅骨和头皮层的电导率值。当磁通密度测量的信噪比（SNR）= 15且目标颅骨与脑电导率比为1/15时，单变量模拟中目标电导率与估计电导率之间的相对误差（RE）为0.0737±0.0746。在三变量模拟中，相对误差为0.1676±0.0317。还通过计算机模拟评估了电极位置不确定性的影响。目前这些有前景的结果表明，从无创磁通密度测量估计头部内重要电导率值是可行的。
