基于FPGA的高频超声成像实验平台

doi:10.3969/j.issn.0258-8021.2015. 05.009

摘要
图/表
参考文献
相关文章 (2)

全文: PDF (11214 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要高频超声具有卓越的分辨能力，在临床和医学基础研究中都得到重要应用。新的图像和信号处理算法不断被提出来，用于提高其成像质量和获取新的信息。为了实验这些算法，需要一个通过编程来改变功能的实验平台，这个平台必须能进行高频超声扫描以采集图像，具备足够实验算法运行的硬件资源，而且能提供方便的调试手段。针对上述应用,提出一种基于FPGA的实验平台，用于研究10～50 MHz高频超声的图像和信号处理的实时算法。该平台具有与实验算法模块相独立的扫描控制器，可以保证在实验算法的调试过程中图像采集和显示均不受影响。系统以180 MHz的采样率采集射频信号，以大规模FPGA及其外围电路提供实验算法运行的空间，并预置可配置的带通滤波器、正交解调、降采样、基带信号幅度检测和对数压缩等处理模块。还开发一个RAM观测工具，专门用于读取处理过程中存入片外RAM的图像数据，以检验中间结果的正确性。采用10 MHz的超声探头，对志愿者眼球进行在体扫描实验，分别对平台的数据采集、带通滤波、幅度提取、对数压缩等模块进行验证。结果表明，带通滤波处理能有效滤除射频信号中的高频、低频噪声，在处理时钟选为50 MHz的条件下，对于每条探测深度为64 mm的扫描线，平台从数据采集至IQ降采样输出完毕，总时间为85.34 μs，而幅度检测和对数压缩算法在340 ns中就可以完成。该处理速度完全能够满足成像的实时性要求。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	周峥郑政^#*

关键词 ：高频超声, 图像处理, 实验平台, FPGA

Abstract：High frequency ultrasound imaging has been extensively applied to smallscale tissue observation in clinic and biomedical research area because of its high-resolution. New processing algorithms have ceaselessly been proposed in this area to improve the image quality. A flexible experimental platform which function can be changed by programming is needed to test the new algorithms. This kind of platform should have the ability to adapt different frequencies of the probes. It also needs to provide sufficient resources to host the algorithms under test and of course the convenient debugging tool is another necessity. We recently developed a Field Programmable Gate Array (FPGA) based experimental platform for these purposes. It scans the tissue by single element probe which central frequency is between 10 MHz and 50 MHz. The platform features an independent scanning control unit to ensure that data acquiring process is always right and the resulting realtime B-mode images are displayed on a screen no matter the algorithm under test works well or not. There is a largescale FPGA on board to host the algorithm under test, and many general purpose processing modules, such as bandpass filter, orthogonal demodulation, down sampling, amplitude detecting and logarithmic compression modules are prepared in the FPGA. In addition, we developed a debug tool. With this tool one can read RAMs outside the FPGA at any addresses and move the data into a computer connected to the FPGA by JTAG. The software based on the MATLB can plot the data on the computer screen as a waveform or a B-mode image depends on the dimensionality of the data that are under check. We tested the platform by scanning the volunteers’ eyes with a 10 MHz probe. Results show that the band-pass filter is effective in reducing the RF noise. It is also showed that, in the condition of 50 MHz clock rate, the total time from the data is sampled to demodulated is only 85.34 μs for each A-line of 64 mm deep, while amplitude detection and logarithmic compression take 340 ns only.

Key words： high frequency ultrasound image processing experimental platform field programmable gate array（FPGA）

基金资助:上海市研究生创新基金项目（JWCXSL1302）

引用本文:

周峥郑政^#*. 基于FPGA的高频超声成像实验平台[J]. 中国生物医学工程学报, 2015, 34(5): 581-588.
Zhou Zheng Zheng Zheng^#*. FPGA Based Experimental Platform for High Frequency Ultrasound B-Mode Imaging. journal1, 2015, 34(5): 581-588.

链接本文:

http://cjbme.csbme.org/CN/10.3969/j.issn.0258-8021.2015. 05.009 或 http://cjbme.csbme.org/CN/Y2015/V34/I5/581

［1］MadsenEL, FrankGR, McCormickMM, et al.Anechoic sphere phantoms for estimating 3-D resolutionof veryhighfrequency ultrasound scanners［J］. IEEE Trans Ultrason Ferroelectr Freq Control, 2010, 57(10): 2284-2292.
［2］Misaridis T, Jensen JA. Use of modulated excitation signals inmedical ultrasound.part i: basic concepts and expected benefits［J］. IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, 2005,52(2): 177-191.
［3］Damerjian V, Tankyevych O, Souag N, et al. Speckle characterization methods in ultrasound images-A review［J］. IRBM, 2014, 35(4): 202-213.
［4］Parker KJ, Doyley MM, Rubens DJ. Corrigendum: Imaging the elastic properties of tissue:the 20 year perspective［J］.Physics in Medicine and Biology, 2011,56(1):R1-R29.
［5］Brands PJ, Hoeks APG, Ledoux LAF, et al.A radio frequency domain complex crosscorrelation model to estimate blood flow velocity and tissue motion by means of ultrasound［J］.Ultrasound in Medicine & Biology, 1997, 23(6):911-920.
［6］Duck FA. Nonlinear acoustics in diagnostic ultrasound［J］. Ultrasound in Medicine & Biology, 2002, 28(1):1-18.
［7］Rohling R, Fung W, Lajevardi P. PUPIL: Programmable ultrasound platform and interface library［C］//EllisRE, PetersTM,eds. Medical Image Computing and ComputerAssisted Intervention  MICCAI 2003.Berlin Heidelberg: SpringerVerlag Berlin Heidelberg, 2003:424-431.
［8］Tortoli P, Bassi L, Boni E, et al. ULAOP: An advanced open platform for ultrasound research［J］. IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, 2009, 56(10): 2207-2216.
［9］Qiu Weibao, Yu Yanyan, Tsang FK, et al.An FPGAbased open platformfor ultrasound biomicroscopy［J］. IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, 2012,59(7):1432-1442.
［10］Kolumban G, Krebesz TI, Lau FCM. Theory and application of software defined electronics: Design concepts for the next generation of telecommunications and measurement systems［J］. IEEE Circuits and Systems Magazine, 2012, 12(2): 8-34.
［11］Li Yamin, Chu Wanming. Implementation of single precision floating point square root on FPGAs［J］. The 5th Annual IEEE Symposium on FieldProgrammable Custom Computing Machines. Napa Valley：IEEE, 1997: 226-232.
［12］骆艳卜，张会生，张斌，等. 一种CORDIC算法的FPGA实现［J］. 计算机仿真, 2009, 26(9): 305-307.