医学超声成像中自适应波束形成方法研究

发布时间：2018-05-28 12:57

  本文选题：超声成像 + 自适应波束形成　； 参考：《华中科技大学》2014年博士论文

【摘要】：作为四大医学影像技术之一,医学超声具有安全、便携、实时、低成本等优点,在临床诊断中已得到了广泛应用。在超声成像系统中,波束形成处于核心位置,对成像质量起着决定性作用。其中,延时叠加(delay-and-sum, DAS)波束形成技术能简单而有效地实现超声图像的重建,已在商业产品中得到广泛的应用。但是,其所形成的波束主瓣宽度较宽、旁瓣水平较高,降低了图像空间分辨率和对比度。传统的变迹技术使用一组预先确定的权重(如汉明窗)来降低旁瓣干扰,但会增加主瓣宽度,牺牲了图像分辨率。自适应波束形成技术通过从接收信号中提取特征信息来确定孔径权重,在成像区域中形成更窄主瓣宽度和更低旁瓣级的高质量波束,能够同时增强分辨率和对比度。本论文以相干因子(coherence-based factor)自适应加权和最小方差(minimum variance, MV)波束形成为主要研究对象,分析它们在超声成像中的优点和不足,并提出了相应的改进方法,具体包括： 首先,幅度相干因子法(the coherence factor, CF)和相位/符号相干因子(thephase/sign coherence factor, PCF/SCF)法能以较低的计算复杂度显著抑制旁/栅瓣,减少杂波,但是会降低图像平均亮度、增大斑点方差甚至损失部分纹理信息。本论文使用复解析孔径数据,设计了一种基于相量离散度的相干因子,通过对相干波束和的加权,抑制旁瓣和杂波,减少离轴干扰信号和随机噪声。仿真和真实体模数据的实验结果表明：相比于DAS,该方法能同时提高分辨率和对比度。但是,它也会产生斑点纹理上的负面效果。为了解决该问题,我们通过一个用户定义的参数α来调节离轴干扰抑制和图像纹理信息之间的平衡。结果表明,一个0.9至1之间、稍小于1的α值能获得最佳的成像结果。 其次,MV波束形成方法能显著提高空间分辨率,但是对旁瓣和对比度的改善尚显不足。为此,本文提出将MV波束形成与相位相干成像相结合的改进方法。实验表明,该方法在保持高分辨率的同时,减少了旁瓣和杂波,但会造成图像亮度降低、斑点方差增大等问题。 然后,详细指出了使用相干因子波束形成方法(如CF、PCF/SCF)重建图像所存在的缺陷,包括：整体图像亮度降低,斑点方差增大,高回声反射体周围出现黑洞伪影以及远场点目标亮度被过低估计等。为解决这些问题,本论文从理论上分析了它们产生的原因,进而提出一种新的空时平滑处理技术来修正相干因子类方法。由此所设计的空时平滑相干因子在一个发射脉冲的持续时间上测量所划分重叠子阵列的波束和之间的信号相干性。仿真和真实体模数据的实验结果表明,该方法能明显提高相干因子成像的鲁棒性,减少斑点方差,消除黑洞伪影,具有较强的杂波抑制能力,从而提高图像对比度,增强了图像中囊肿的可检测性。 最后,为了进一步改善空时平滑相干因子方法的成像质量,提出了三种改进方案,即对该相干因子进行空间滤波、将其与空间复合技术相结合以及与最小方差波束形成相结合。仿真实验结果表明：空时平滑相干因子经过空间滤波后能减少由于因子值的波动引起的斑点亮度变化,消除黑洞伪影,进一步增强图像对比度；与空间复合的结合能更好地减少斑点噪声和杂波,使背景更加均匀,显著提高对比度；与MV波束形成相结合,可同时增强空间分辨率和对比度,并且较好地保持了背景的纹理信息,获得整体图像质量的提升。 综上所述,本论文通过对相干因子和MV波束形成方法的改进,弥补了它们在超声成像中的不足,提高了其成像性能,使之更适合临床应用。尤其是,空时平滑相干因子自适应加权方法的提出,为提高超声成像质量、增强超声系统的检测诊断能力提供了有益的尝试。
[Abstract]:As one of the four major medical imaging techniques, medical ultrasound has the advantages of safety, portability, real time, low cost and so on. It has been widely used in clinical diagnosis. In the ultrasonic imaging system, the beam forming is at the core position and plays a decisive role in the imaging quality. Among them, the delay-and-sum, DAS beam forming technology can be simple. The effective realization of the reconstruction of ultrasonic images has been widely used in commercial products. However, the width of the main lobe of the beam is wide, the side lobe level is high, and the spatial resolution and contrast of the image are reduced. The traditional technique of tracking uses a set of predetermined weights (such as Han Mingchuang) to reduce the sidelobe interference, but it will increase the main point. The adaptive beamforming technology determines the aperture weight by extracting feature information from the received signal to form a higher quality beam with narrower main lobe width and lower sidelobe level in the imaging region, which can enhance the resolution and contrast at the same time. This paper uses the coherent factor (coherence-based factor) self adaptation in this paper. Minimum variance (MV) beamforming is the main research object, and the advantages and disadvantages of them in ultrasonic imaging are analyzed, and the corresponding improvement methods are put forward, which include:
First, the the coherence factor (CF) and the phase / symbol coherent factor (thephase/sign coherence factor, PCF/SCF) can significantly reduce the side / gate flap with a lower computational complexity, reduce the clutter, but reduce the average image brightness, increase the spot variance and even lose some texture information. This paper uses complex analysis. In aperture data, a coherent factor based on phasor dispersion is designed. By weighting the coherent beam and reducing the sidelobe and clutter, the off axis interference signal and random noise are reduced. The experimental results of the simulation and real model data show that the method can increase the resolution and contrast at the same time compared to the DAS. However, it also produces spots. In order to solve this problem, we use a user defined parameter alpha to adjust the balance between off axis interference suppression and image texture information. The results show that an alpha value of less than 1 between 0.9 and 1 can obtain the best imaging results.
Secondly, the MV beamforming method can significantly improve the spatial resolution, but the improvement of the sidelobe and contrast is still inadequate. Therefore, this paper proposes an improved method combining the MV beam forming with phase coherent imaging. The experiment shows that the method reduces the sidelobe and clutter while maintaining high resolution, but reduces the brightness of the image, The enlargement of the speckle variance.
Then, the defects in the image reconstruction using the coherent factor beamforming (such as CF, PCF/SCF) are described in detail, including the reduction of the brightness of the whole image, the increase of the speckle variance, the black hole artifact around the high echo reflector and the low estimation of the brightness of the far point target. In order to solve these problems, the paper has theoretically analyzed the problem A new space time smoothing technique is proposed to modify the coherent factor class method. The designed space-time smooth coherence factor is used to measure the beam and signal coherence between the overlapped subarrays on the duration of a emission pulse. This method can obviously improve the robustness of coherent factor imaging, reduce the speckle variance, eliminate the black hole artifact, and have strong clutter suppression ability, thus improving the contrast of images and enhancing the detectability of the cysts in the image.
Finally, in order to further improve the imaging quality of the space-time smoothing coherent factor method, three improvements are proposed, namely, the spatial filtering of the coherent factor, combining it with the spatial composite technology and combining with the minimum variance beamforming. The simulation results show that the space time smoothing coherent factor can be reduced after spatial filtering. Less speckle brightness changes caused by the fluctuation of the factor value, eliminate the black hole artifact and further enhance the contrast degree of the image. The combination with the space compound can better reduce the speckle noise and clutter, make the background more uniform and improve the contrast significantly; combined with the MV beam formation, the spatial resolution and contrast can be enhanced at the same time, and better, and better The background texture information is maintained and the overall image quality is improved.
To sum up, by improving the coherent factors and MV beamforming methods, this paper makes up their shortcomings in ultrasonic imaging, improves their imaging performance and makes it more suitable for clinical applications. Especially, the adaptive weighting method of space-time smooth coherence factor is proposed to improve the quality of ultrasound imaging and enhance the detection and diagnosis of ultrasonic systems. Ability provides a useful attempt.
【学位授予单位】：华中科技大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：R445.1

【参考文献】

中国博士学位论文全文数据库 前1条

1 李雅琴;非衍射超声声源的仿真与实现研究[D];华中科技大学;2012年

，

本文编号：1946781