热声效应对磁声成像中组织电导率图像重建影响研究

Images of electrical characteristics to reveal early phase of malignant tumor can be obtained by Magneto-acoustic Imaging that is promising in early diagnosis. The current paradigm is to generate sound vibration based on magneto-acoustic coupling effect within a pulse magnetic field and a steady magnetic field, and then to record and process the acoustic source that is the divergence of Lorentz force density, and finally reconstruct the conductivity images derived from the sound signal. In fact, there is not only vibration caused by Lorentz force within steady magnetic field, and also vibration without steady magnetic field which also known as Thermos-acoustic effect, which is more obvious in biological tissues. Besides, the acoustic source of Thermos-acoustic effect has the frequency component as that of the magneto-acoustic effect, and both of them are related to the electrical characteristics of biological tissues. We assume that the sound signal received by transducers is the superposition of these two effects, and thus, the de facto received signal used in reconstruction is not just the sound characteristics of sound pressure caused by Lorentz force used in solving the wave equations. Therefore, the conductivity image restricted is inaccurate. We propose two routes to study the mechanism of influence of Thermos-acoustic effect on reconstruction of conductivity image in Magneto-acoustic imaging: 1) to propose a novel algorithm to reconstruct the conductivity images using the sound source of Lorentz force superimposed by that of Thermos-acoustic effect; 2) to separate the sound source of Lorentz force by filtering that of Thermos-acoustic effect, and to reconstruct the conductivity image using the original algorithm. We believe that in this way we can reduce the error raised by assuming that the sound source is solely caused by Lorentz force, obtain the characteristics of these two sound sources, improve the detection method of tissue conductivity, propose a novel algorithm of image construction, and improve the quality of conductivity image reconstruction.

磁声成像方法可以获取反映恶性肿瘤早期病变的电特性图像，旨在实现早期诊断。该方法研究在脉冲磁场和稳恒磁场共同作用下基于磁声耦合效应产生的声振动，以洛伦兹力密度散度声源理论进行声信号特征提取及电导率图像重建。实际上，磁声成像中不仅存在洛伦兹力引起的声振动，也有无稳恒磁场的声振动（即热声效应），在生物组织内尤其明显，且与组织电特性相关。由于声传感器接收到的声信号是两种效应声信号的叠加，与目前声信号求解基于洛伦兹力密度散度为唯一声源建立的声压波动方程存在差异，导致重建电导率不准确。本项目拟通过两条途径研究热声效应对电导率图像重建的影响：1）建立两种效应叠加声源的电导率重建算法；2）滤除热声效应信号，分离出单一洛伦兹力声信号，用原有声源理论重建电导率。进而减小只考虑洛伦兹力声源对声信号解析造成的误差，获得真实的声信号特征，改进组织电导率信息的检测方法，研制优化的图像重建算法，提高电导率成像质量。

1）磁声成像中热声效应的发现与验证。通过对不同材料待测样本有无静磁场情况下的声信号的时频分析，发现了在低电导率类生物组织及生物组织材料中占优的非磁声信号不可忽略；进一步对该磁声信号的声源项进行实验研究，验证该非磁声声源为与激励电场二次相关的标量源，结合磁声成像系统结构特点，判断该非磁声信号为组织热声效应产生的热声信号，即磁声成像中存在着热声效应的影响。.2）考虑热声效应的新的声源理论模型的建立.考虑磁声成像过程中同时发生的磁声效应与热声效应引起的声振动，明晰磁声源与热声源声源发生的物理过程和声源机制，提出新的磁声+热声混合声源的新声源理论模型，在此基础上建立新的磁热声声源理论方程。.3）基于新声源理论的仿真研究.基于已建立的磁声+热声混合声源的新理论模型，建立新的磁声成像算法，使用椭球仿真模型对单一声源和混合声源的重建分布进行比较研究；并且，考虑真实的实验条件，分别对几种不同材料样本的磁声信号、热声信号及磁热声混合信号进行了理论仿真，验证了新的声源理论的可行性。.4）磁声实验系统的改进优化.对磁声成像实验系统进行优化：改进静磁铁装载模式，使其可以实现磁声成像模式与热声成像模式的灵活切换；采用电控精密位移导轨代替之前的手动位移导轨，提高样本及探头位置的准确性及稳定性。采用编码激励的方法在提高实验系统的信噪比达到80dB信号，实现对声压信号检测灵敏度达到10-6Pa。.5）基于新声源理论的磁声实验研究.采用圆环形导电橡胶圈作为实验样本，分别对其进行磁声成像与热声成像实验，从实测样本信号中估算热声信号约占混合声信号的1/4，通过实验信号验证了新声源理论。基于新的声源理论和热声声源理论分别进行图像重建，分析比较不同声源重建图像的特点，进一步验证了新的声源理论及重建算法的正确性。并且，首次使用人体手指进行了在体的磁声成像实验，在体检测到了手指上的磁声信号。