中文摘要：

目的 采用电阻抗成像技术重建三维头部模型的阻抗分布图像，检测组织是否发生病变。方法 在有限元球头模型和真实头部模型两种三维组织模型上进行仿真，采用微分进化算法重构组织图像，有效定位阻抗突变区域，检测病变组织部位。结果 该算法能够精确重建组织图像，成功检测病变区域。结论 本法是一种简单、鲁棒性强的进化类全局优化算法，用于电阻抗成像技术中，进化总能得到很好收敛，成像质量较高，可靠性较强。

英文摘要：

Objective To reconstruct the real conductivity images of 3D head model, and to detect brain lesion by means of electrical impedance tomography （EIT） technology. Methods In order to locate the pathological changes of head impedance effectively, a series of computer simulation were conducted on a finite element model and a realistic-geometry head model, and differential evolution algorithm was adopted to reconstruct the conductivity image of the head tissue. Results The proposed algorithm could accurately reconstruct the impedance of the head tissue and successfully detect lesion area presently. Conclusion The global optimization and evolution algorithm can be applied to EIT simply with good convergent and robustness. The reconstructed images show higher quality and are reliable.