Research On The Algorithm Of Microwave Induced Thermos-acoustic Tomography In Heterogeneous Environment

Microwave induced thermo-acoustic tomography(MITAT)combines the advantages of both microwave imaging and ultrasound imaging to make its inversion images have high contrast and high resolution,making it an extremely useful application in prospective biomedical detection technology.Due to its multi-physical properties,the corresponding inversion imaging process is very complicated.And the inversion process is an inverse scattering process,in which the inhomogeneity distribution of each parameter involved will affect the inversion imaging results.At present,the correlation research of many parameters is still in progress.It is difficult to control the variable to blame the influence of each parameter on the inversion result on one or two parameters.By analyzing the existing microwave thermal-induced ultrasound theory,it is found that the acoustic velocity distribution has a very important influence on the size,position and shape of the inversion target.This paper is aimed at the qualitative inversion of the target,so we made a certain choice in the selection of environmental parameters,selected the acoustic velocity as the main research object,and studied the microwave thermal-induced thermoacoustic imaging algorithm under the inhomogeneity distribution of acoustic velocity.The main research content of this paper can be divided into the following parts:1.For the problem of extracting thermoacoustic signals under low signal-to-noise ratio in microwave-induced thermoacoustic imaging,in this thesis proposes a noise reduction filtering algorithm based on wavelet transform and singular value decomposition.Based on the time-frequency characteristics of the thermoacoustic signal and the existing microwave thermoacoustic system noise signal of the research group,the noise signal is filtered in wavelet domain and singular value space,then weak thermoacoustic signals submerged in ambient noise are extracted.The difficulty in extracting thermoacoustic signals under low signal-to-noise ratio in the process of microwave thermally induced thermoacoustic imaging is solved,which provides an important premise for the accuracy of subsequent inversion imaging.2.In order to estimate the velocity distribution in inhomogeneous environments,a priori self-focusing algorithm is proposed,which combines the accuracy of a priori information method with the stability of a self-focusing algorithm.Without increasing the complexity of the experimental system and the computational complexity,a more accurate estimation of the sound velocity distribution is obtained,which provides an important guarantee for improving the quality of inversion imaging3.For the low image quality caused by inhomogeneous parameters in the environment,if this thesis does not enhance the image features of the inversion,it will affect the diagnosis of the nature of the target.By combining the time dimension and space dimension of the inversion results,the method introduces the acoustic velocity difference term,and the environment background noise and artifacts are suppressed while enhancing the target characteristics.So this method can reduce the possibility of misjudgment in disease diagnosis4.In this thesis,a dictionary reconstruction method based on inhomogeneity acoustic speed is proposed to reduce the influence of inhomogeneity acoustic speed and thermoacoustic attenuation.The compression sensing microwave thermoacoustic imaging algorithm is based on the model reconstruction method,which has a very significant advantage in thermal acoustic attenuation suppression.This method needs to establish a thermoacoustic dictionary first,and the existing related algorithms are all based on a single homogeneity acoustic speed model.After solving the optimization process,the image will eventually be distorted.By introducing the non-uniform distribution of sound velocity to the existing dictionary building method,the model matrix is more accurate,which reduces the effects of inhomogeneity acoustic velocity and thermoacoustic attenuation,and better reconstructs the target.