Zlochiver Sharon, Rosenfeld Moshe, Abboud Shimon
Biomedical Engineering Department, Tel-Aviv University, Tel-Aviv 69978, Israel.
IEEE Trans Med Imaging. 2003 Dec;22(12):1550-60. doi: 10.1109/TMI.2003.820025.
A reconstruction algorithm, based on the modified Newton-Raphson algorithm, was developed for induced-current electrical impedance tomography and studied in theoretical two-dimensional geometry representing a human thorax. The finite-volume method was applied for the discretization of the physical domain, resulting in a symbolic representation of the Jacobian matrix, which is accurate and fast to construct. Several system configurations, differing in the number of excitation coils and electrodes, were simulated, and the performance in thoracic imaging was studied. It was found that a six-coil system shows a significant 40% improvement of conductivity values reconstruction over the three-coil system (an error of 2.06 omega(-1) compared with 3.44 omega(-1)). A number of 32 electrodes was found to be sufficient, being the smallest number of electrodes to still provide a reasonable performance (only 4.2% degradation in average conductivity error compared with the maximum possible 106-electrode system).
基于改进的牛顿-拉夫逊算法，开发了一种用于感应电流电阻抗断层成像的重建算法，并在代表人体胸部的理论二维几何模型中进行了研究。采用有限体积法对物理域进行离散化，得到雅可比矩阵的符号表示，其构建准确且快速。模拟了几种不同激励线圈和电极数量的系统配置，并研究了其在胸部成像中的性能。结果发现，六线圈系统在电导率值重建方面比三线圈系统有显著的40%的提升（误差为2.06Ω⁻¹，而三线圈系统为3.44Ω⁻¹）。研究发现32个电极就足够了，这是仍能提供合理性能的最小电极数量（与最大可能的106电极系统相比，平均电导率误差仅下降4.2%）。
