CORRECTION OF ULTRASONIC WAVE-FRONT DISTORTION USING BACKPROPAGATION AND A REFERENCE WAVE-FORM METHOD FOR TIME-SHIFT COMPENSATION

CORRECTION OF ULTRASONIC WAVE-FRONT DISTORTION USING BACKPROPAGATION AND A REFERENCE WAVE-FORM METHOD FOR TIME-SHIFT COMPENSATION
复制标题

DOI:
10.1121/1.410304
复制
发表时间:
1994-08-01
影响因子:
2.4
通讯作者:
WAAG, RC
WAAG, RC
中科院分区:
物理与天体物理3区
文献类型:
--
作者:
LIU, DL;WAAG, RC

文献摘要

被引文献

相似文献

介绍了超声脉冲在人体腹壁试样中传播时产生的振幅、形状畸变和到达时间波动的模型。在该模型中,当波前通过只引起时移的相位屏后在均匀介质中传播时,振幅和形状会发生畸变。利用角谱法对波前进行反向传播来补偿这种畸变。补偿采用由点状源发射的波形,并在通过组织传播后进行测量。首先对波形进行几何路径校正,然后在一系列不断增加的距离上反向传播。在每个距离上,计算波形相似系数,以找到波形最相似的反向传播距离。设计了一种新的脉冲到达时间估计方法,用于几何校正和时移补偿。该方法自适应导出参考波形,然后与孔径内的所有波形交叉相关,从而获得到达时间的曲面。对曲面进行迭代平滑,去除因波形畸变引起的离群点。14个试样的波形相似系数初始均值(+/-s.d.)为0.938(+/-0.025)。将波形反向传播至各试件最大波形相似距离后,波形相似系数提高至0.967(+/-0.015)。波前相应的能级波动初始为4.2 (+/-0.4)dB,反向传播后变为3.3 (+/-0.3)dB。对于聚焦于180 mm的波前,在孔径中进行时移补偿时,-30 dB平均(+/-s.d)有效聚焦半径为4.2 (+/-1.2)mm,在反向传播后进行时移补偿时变为2.5 (+/-0.5)mm。这些结果表明,在离孔径一定距离处放置相位屏是描述人体腹壁产生的波前畸变的改进模型,波前反向传播后进行时移估计和补偿是补偿这种畸变的有效方法。
A model is introduced to describe ultrasonic pulse amplitude and shape distortion as well as arrival time fluctuation produced by propagation through specimens of human abdominal wall. In the model, amplitude and shape distortion develops as the wavefront propagates in a uniform medium after passing through a phase screen that only causes time shifts. This distortion is compensated by a backpropagation of the wavefront using the angular spectrum method. The compensation employed waveforms emitted by a pointlike source and measured after propagation through the tissue. The waveforms were first corrected for geometric path and then were backpropagated over a sequence of increasing distances. At each distance, a waveform similarity factor was calculated to find the backpropagation distance at which the waveforms were most similar. A new method was devised to estimate pulse arrival time for geometric correction as well as to perform time-shift compensation. The method adaptively derives a reference waveform that is then cross correlated with all the waveforms in the aperture to obtain a surface of arrival times. The surface was smoothed iteratively to remove outlying points due to waveform distortion. The mean (+/-s.d.) of the waveform similarity factor for 14 specimens was found to be 0.938 (+/-0.025) initially. After backpropagation of waveforms to the distance of maximum waveform similarity for each specimen, the waveform similarity factor improved to 0.967 (+/-0.015). The corresponding energy level fluctuation in the wavefront was 4.2 (+/-0.4) dB initially and became 3.3 (+/-0.3) dB after backpropagation. For wavefronts focused at 180 mm, the -30 dB mean (+/-s.d.) effective radius of the focus was 4.2 (+/-1.2) mm with time-shift compensation in the aperture and became 2.5 (+/-0.5) mm with backpropagation followed by time-shift compensation. These results indicate that a phase screen placed some distance away from the aperture is an improved model for the description of wavefront distortion produced by human abdominal wall and that wavefront backpropagation followed by time-shift estimation and compensation is an effective method to compensate for such distortion.