高分辨率超声衍射层析成像重建算法的研究

项目名称： 高分辨率超声衍射层析成像重建算法的研究

项目编号： No.61302179

项目类型： 青年科学基金项目

立项/批准年度： 2014

项目学科： 无线电电子学、电信技术

项目作者： 方杰

作者单位： 皖西学院

项目金额： 23万元

中文摘要： 超声衍射层析成像是一种定量的新的成像技术，从安全和价格角度考虑，它优于X射线等带有电离辐射的层析成像技术。衍射层析成像在复平面上呈不同半径圆弧状分布，显然，经典的层析成像的重建算法不能简单地应用于多频率超声衍射层析成像，因此，本项目拟以多频率超声波衍射层析成像及其重建算法为主要研究内容，利用近期出现压缩感知理论可以有效地解决采样数据量这一突出问题，设计高频超声压缩采样系统，可获得比奈奎斯特采样小得多的传输数据量。探讨接收端数据实现高概率精确重建方法，为此目标我们必须解决基于非均匀快速傅里叶变换卷积核的选取问题、快速密度补偿及网格化问题以及超声波散射场正则化模型建立问题。

中文关键词： 衍射层析成像；密度补偿；压缩感知；非均匀傅里叶变换；总体变差

英文摘要： Unlike conventional ultrasound imaging technique，ultrasound tomography is a novel and quantitative one , and is superior to X and other ionizing radiation tomography considering security and expense. What's more, the Fourier transform of projection data from multi-band freuency ultrasonic is distributed over complex plane in form of different radius arcs. Obviously, the mathematical basis used for image reconstruction in straight ray tomography is not suitable for it. The main research works and innovations in this project are the multi-band ultrasound diffraction tomography and its reconstruction algorithms which will generate a tremendous amount of data and inflict great pressure on signal acquisition and transmission ，these problems can be efficiently solved by Compressive sense theory --through designing compressed sampling system of high frequency ultrasonic ，the sampling data that is much smaller than that of Nyquist sampling is received at terminal。 Investigate the use of data to achieve a high probability of accurate reconstruction method of the receiving end. An accurate reconstruction at high probability at the reception terminal can be obtained，we must resolve the Correct selection of the convolulion kernel based on non-uniform fast Fourier transform ,improve the efficiency of density compensation

英文关键词： diffraction tomography；density compensation；compressive sensing；non-uniform FFT；total variation