基于微波热声层析成像定量重建生物组织电导率的改进方法

An Improved Method for Quantitative Reconstruction of Biological Tissue Conductivity Based on Microwave-induced Thermoacoustic Tomography

目的生物电磁学参数中的电导率与组织的功能性信息直接相关,精准重建生物组织电导率在医学成像技术和医学诊断领域中有着重要意义。本文改进定量微波热声层析成像(microwave-induced thermoacoustic tomography,MTAT)算法,使组织电导率的重建精度提高。方法本文在利用有限元离散法求解热声波动方程和亥姆霍兹方程的基础之上,提出了一种基于正则化牛顿迭代法(regularized Newton iteration method,RNIM)定量重建组织电导率的改进方法。结果通过数值模拟实验和含不同浓度NaCl溶液的仿体实验,验证了算法改进的有效性。组织仿体实验结果表明,目标在不同位置、不同大小、不同对比度情况下,相比于定量微波热声层析成像采用拟合(fitting)的方法,采用正则化牛顿法定量重建的仿体电导率相对误差明显降低,重建目标精度提高。在仿体实验中采用RNIM方法重建相同浓度的单目标在不同位置的电导率变化幅度更小,以及重建多目标电导率的相对比值与实际更接近,实验结果验证了改进方法的稳定性。结论研究结果表明优化算法能更加准确地定量重建组织仿体的电导率,这对于肿瘤的定位和分期的早期筛查及精准诊疗,预防疾病恶化具有重要意义。

Objective The conductivity in bioelectromagnetic parameters is directly related to the functional information of the tissue, and precise reconstruction of biological tissue conductivity is of great significance in the fields of medical imaging technology and medical diagnosis. In this paper, the microwave-induced thermoacoustic tomography(MTAT) algorithm is improved to improve the reconstruction accuracy of tissue conductivity.Methods On the basis of utilizing the finite element discrete method to solve thermoacoustic wave equation and Helmholtz equation, an improved method of quantitative reconstruction of biological tissue conductivity based on regularized Newton iteration method(RNIM) is proposed in this paper. Results The effectiveness of the algorithm improvement was verified by numerical simulation and phantom experiments with different concentrations of NaCl solution. The results of the tissue phantom experiments showed that the relative error of the phantom conductivity quantitatively reconstructed by the regularized Newton method is significantly lower than that of the quantitative MTAT with fitting, for the cases with different target positions, sizes and contrasts.And the accuracy of the reconstruction is improved. Simultaneously, the RNIM method was used to reconstruct the conductivity of a single target with the same concentration at different positions with smaller variation in the mimic experiments, as well as the relative ratio of the reconstructed conductivity of multiple targets was closer to the actual one, which the experimental results verified the stability of the improved method. Conclusion The results show that the optimized algorithm reconstruct the conductivity of the tissue phantom more accurately and quantitatively, which is of great significance for the early screening and precise diagnosis of tumor localization and staging, to prevent the deterioration of the disease.