High-resolution ultrasound displacement measurement apparatus and method

[0001] None

[0002] Not Applicable.

[0003] Not Applicable.

[0004] The present invention relates to non-invasive measurements of displacement, velocity, thickness, and strain in both medical and industrial applications.

[0005] In traditional pulse echo techniques for the measurement of displacement, ultrasound (radar, or sonar) waves are transmitted and the echoes from the object are received. The displacement is measured by comparing the time shift, frequency shift or phase shift between two echoes. For example, traditional ultrasound methods to measure the thickness of the cornea (about 500 nm) is to derive it from the time interval between two peaks of the echoes from the two sides of cornea. The traditional ultrasound method to measure the velocity of the blood flow is to measure the Doppler frequency shift in time domain or frequency domain. The traditional ultrasound imaging or measurement of elasticity derives the displacement or strain within a body by detecting the time shift of the echoes through correlations in time domain or frequency domain.

[0006] The resolution of the displacement measurement based on traditional pulse-echo technique is limited by the center frequency of the transmitted wave, the sampling rate of the echoes, and electronic noise in the measurement system. There is a need in the art to overcome these limitations and to increase accuracy, resolution, responsiveness, sensitivity, resistance to noise and speed.

[0007] Further areas of applicability of the present invention will become apparent from the detailed description provided hereinafter. It should be understood that the detailed description and specific examples, while indicating the preferred embodiment of the invention, are intended for purposes of illustration only and are not intended to limit the scope of the invention.

[0008] In general, the invention provides a system for determining the displacement of an object or a desired region in an object with high resolution and the measurement is highly robust to noise. The system includes transmitting a pair of Golay's complementary sequences (GCS), receiving two echoes from the same object or the same region of an object, compressing the pulse, and eliminating the mainlobe. The displacement of the object between the two transmissions is then derived from the residual clutter signals around the mainlobe of the compressed pulse output.

[0009] Implementations of the invention may include one or more of the following features. A signal generator circuit generates the Golay's complementary sequences to control the device of wave transmitter. The wave transmitter can be an ultrasound transducer, a radar antenna, or a sonar transmitter. Echoes from the targeted region of the object are then received by a receiving device. In some situations, the device of wave transmitter and the receiving device can be the same device. A signal conditioning circuit performs pre-amplification, compensation for attenuation of the echoes, and band-pass filtering of the echoes. An analog-to-digital converter (ADC) samples the analog echoes into digital echoes. A digital signal processing device or a computer performs the pulse compression, mainlobe elimination, clutter collection, and displacement calculation (FIG. 1).

[0010] In another aspect, the invention provides a system for determining the velocity of an object or a desired region in an object. The system includes transmission of a pair of GCS sequences (A and B). The two corresponding echoes from the same object or the same region of an object are then received. The pulse is compressed and the mainlobe is eliminated. The displacement of the object or the desired region between two transmissions is derived from the residual clutter signals around the mainlobe of the compressed pulse output. The velocity of the object or the desired region is then derived from the displacement and time interval between the two transmissions.

[0011] In another aspect, the invention provides a system for determining the thickness of an object or a desired region in an object. The system includes transmission of a pair of GCS sequences. The two corresponding echoes from two sides of the same object or the same region are then received, with the pulse compressed and the mainlobe eliminated subsequently. The thickness of the object or the desired region is then derived from the residual clutter signals around the mainlobe of the compressed pulse output.

[0012] In another aspect, the invention provides a system for determining the strain in an object or a desired region in an object with high resolution and the measurement is highly robust. The system includes transmission of a pair of Golay's complementary sequences. The two corresponding echoes from the same object or the same region of an object are then received, with the pulse compressed and the mainlobe eliminated subsequently. The displacement of the areas or regions of an object between the two Golay transmissions is derived from the residual clutter signals around the mainlobe of the compressed pulse output. The strain of the object or the desired region is then derived from the displacements and overall size of the object or the desired region.

[0013] In another aspect, the invention provides a system for 2D or 3D mapping of the displacement in an object or a desired region in an object with high resolution and robust measurement. The system includes transmitting of a pair of Golay's complementary sequences. The two corresponding echoes from the same object or the same region of an object are then received, and the pulse is compressed and the mainlobe is eliminated subsequently. The displacement of the regions of the object between two transmissions is derived from the residual clutter signals around the mainlobe of the compressed pulse output. The 2D or 3D mapping of the displacement is then derived from the displacements measured in different positions in the 2D or 3D space.

[0014] In another aspect, the invention provides a system for 2D or 3D mapping the velocity of an object or a desired region in an object. The system includes transmitting of a pair of GCS sequences. The two corresponding echoes from the same object or the same region are then received, with the pulse compressed and the mainlobe eliminated subsequently. The displacement of the object or the desired region between two transmissions is then derived from the residual clutter signals around the mainlobe of the compressed pulse output. The velocity of the object or the desired region is then derived from the displacement and the time interval between the two transmissions. The 2D or 3D mapping of the velocity is then derived from velocities measured in different positions in the 2D or 3D spaces.

[0015] In another aspect, the invention provides a system for 2D or 3D mapping of the strain of an object or a desired region in an object with high resolution and robust measurement. The system includes transmitting of a pair of Golay's complementary sequences. The two corresponding echoes from the same object or the same region of an object are then received, pulse compressed and mainlobe eliminated. The displacement of the scatterers between two transmissions is derived from the residual clutter signals around the mainlobe of the compressed pulse output. The strain of the object or the desired region is then derived from the displacement and length of the object or the desired region.

[0016] A digital signal processing device or a computer performs pulse compression, mainlobe elimination, clutter collection, and displacement calculation. The 2D or 3D mapping of the strain is then derived from the strain measured in different positions in the 2D or 3D space.

[0017] Various aspects of the invention may provide one or more of the following advantages. Mechanical properties of an object or a region in an object, such as elastic stiffness and viscous damping, can be derived from the displacement, velocity, thickness, strain of the object or regions in the object.

[0018] Those and other advantages of the invention, along with the invention itself, will be more fully understood after a review of the following figures, detailed description and claims.

[0036] The following description of the preferred embodiment(s) is merely exemplary in nature and is in no way intended to limit the invention, its application, or uses.

[0037] Embodiments of the invention provide techniques for non-invasively measuring displacement, velocity, thickness, and strain in medical and industrial applications. For example, systems of the invention can detect internal displacement or strain in vivo under internal or external perturbations, which is related to the tissue elastic modulus or stiffness. To measure the displacement, an ultrasound transducer is excited by a pair of complementary sequences. In the depicted embodiment, these are Golay Complementary Sequences (GCS) with a controlled time interval related to the distance of the tissue to be measured and the speed of the ultrasound in the tissue. The two echoes from the tissue are then picked up by the same or another ultrasound transducer. In some situation, the echoes need to be pre-amplified to compensate for the energy loss in the tissue. Then, the analog echoes are sampled by Analog-to-digital converter and transferred to a computer or an electronic device (a microprocessor, a digital signal processor, or FPGA). The computer performs the operation of pulse compression, mainlobe elimination, clutter collection, and displacement calculation.

[0038] Referring to FIG. 1, a system for displacement measurement 10 includes the target to be measured 11, wave transmitter and echo receiving devices (transducers) 12, signal generator for coded-excitation 13, pre-amplifier, filter and time gain compensation (TGC) device 14, analog/digital converter ADC 15, PC computer or microprocessor 16. The set 14 may have a functional block for timing control of the coded excitation, TGC and ADC 17, pulse compression of the echoes A and B 18, mainlobe elimination 19, clutter collection 20 and displacement detection 21.

[0039] The wave transmitter and echo receiving devices (transducers) 12 are the transducers for generating the ultrasound (or radar, sonar) wave toward the target and receiving the echoes from the target. The signal generator for the coded excitation 13 is configured to generate the GCS pair to control the transducer. The pre-amplifier, filter and TGC device 14 is configured to pre-amplify, filter and compensate for the energy loss of the echoes caused by attenuation. The ADC 15 is configured to convert the analog echoes to digital echoes. The PC computer or microprocessor 16 is configured to control the timing of the coded excitation, TGC and ADC 17, performing pulse compression of the echoes A and B 18, performing mainlobe elimination 19, performing clutter collection 20, and performing displacement detection 21.

[0040] The Golay's complementary sequences are a pair of sequences A (a₀, a₁, . . . , a_N-1) and B (b₀, b₁, . . . , b_N-1). The length of A and B are N. The element of A and B are either −1 or 1. In the depicted embodiment, a feature of the GCS is:
c_j+d_j=2N j=0
c_j+d_j=0 j±0  (1)
where c_j and d_j are the auto-correlation of GCS A and GCS B:

[0041] $\begin{array}{cc}{c}_j=\{\begin{array}{cc}\sum _{i=j}^{i=N-1}{a}_i{a}_{i-j}& j=0,\dots ,N-1\\ \sum _{i=0}^{i=N-1+j}{a}_i{a}_{i-j}& j=-N+1,\dots ,-1\end{array}& \left(2\right)\\ d_j=\{\begin{array}{cc}\sum _{i=j}^{i=N-1}{b}_i{b}_{i-j}& j=0,\dots ,N-1\\ \sum _{i=0}^{i=N-1+j}{b}_i{b}_{i-j}& j=-N+1,\dots ,-1\end{array}& \left(3\right)\end{array}$

[0042] The pulse compression is a digital signal processing process to obtain sharp echoes from the target. When the GCS A and GCS B are used to control the transducer sequentially, and then the auto-correlations of the two echoes from the same object are summed, a very sharp pulse called compressed pulse or mainlobe will be obtained (demonstrated in FIG. 2). The operation of the auto-correlation and summation to obtain the compressed pulse is called pulse compression.

[0043] A very sharp echo (also called mainlobe) can be obtained from the pulse compression. The displacement of the object during the time interval between GCS A and GCS B induces noise (also called clutter) around the mainlobe and reduction of the mainlobe amplitude.

[0044] FIG. 3 shows the effect of the displacement of the target on the compressed pulse. FIG. 3(a) shows the compressed pulse (summation of the two auto-correlations) when the object does not move between GCS A and GCS B. In FIG. 3(b), (c) and (d), the object moves 7.5 μm, 15 μm and 30 μm between GCS A and GCS B, respectively. In the simulation the center frequency f₀ and −3 dB bandwidth η of the transducer used are 10 MHz and 70%, respectively. The length N and width T_c of the GCS are 256 and 100 ns respectively. The sampling rate f_s of the echoes is 100 MHz.

[0045] In 2-D ultrasound imaging, the clutters cause artifacts in the image. FIG. 4 shows the effect of the tissue movement in the ultrasound images without logarithmic compression. In FIG. 4(a), the tissue did not move between GCS A and GCS B. FIGS. 4(b), (c) and (d) show the amplitude of summed auto-correlations of the echoes when the tissue moved 7.5 μm, 15 μm and 30 μm between GCS A and GCS B, respectively. The image size is about 5 mm×5 mm.

[0046] FIG. 5 shows the relationship between the object movement between the GCS A and GCS B stimulations and the signal to clutter ratio (SCR) of the compressed pulse. The movement of the object between GCS A and B stimulations is related to the SCR of the compressed pulse. The center frequency f₀ and −3 dB bandwidth η of the transducer used simulation are 10 MHz and 70%, respectively. The length N and width T_c of the GCS are 256 and 100 ns respectively. The sampling rate f_s of the echoes is 100 MHz.

[0047] When the two echoes (from GCS A and GCS B respectively) have some time shift, the sum of the auto-correlations of the two echoes will produce clutters. The amplitude of the clutters is proportional to the time shift. The larger the time shift is, the larger the clutter. In FIG. 6(a) partial waveforms of the auto-correlations of the echoes from GCS A and GCS B are shown by the solid and dotted curves, respectively. The dotted waveform is shifted by 5/10 of the width of the GCS code. FIG. 6(b) shows clutters produced by different amounts of time shifts ( 1/10, 2/10, 3/10, 4/10 and 5/10 of the width of the. The length N of the GCS is 256. The X-axis is the sampling point n of the auto-correlation function.

[0048] In traditional pulse-echo methods, the displacement is derived from the time shift between two echoes. The accuracy of the calculation of the time shift is limited by the noise level, bandwidth of the transducer and the sampling rate of the ADC.

[0049] In this invention, the amplitude of the clutter (treated as noise in traditional methods), instead of the time shift, is used to derive the displacement. The bandwidth of the transducer and the sampling rate of the ADC have no direct effect on the amplitude of the clutter. The noise has little effect on the clutter since the pulse compression eliminates most of the white noise. Therefore, the amplitude of the clutter is much more robust measurement of the object displacement than the time shift.

[0050] The displacement of the object has a relationship with the SCR (signal to clutter ratio) of the compressed pulse with the Golay coded excitation (see FIG. 5). However, the SCR is not a reliable measure of the object displacement due to the limitation of the sampling precision and the noise in a physical system. Therefore, a new parameter S_ac is used to replace the SCR to measure the object displacement. S_ac is the sum of the absolute value of all clutters. Usually, the width of the mainlobe equals to the width of one single pulse. Therefore,

[0051] $S_{\mathrm{ac}}=\sum _{t=0}^{t=L/2-l}P_c\left(t\right)+\sum _{t=L/2+l}^{t=L}P_c\left(t\right)$
where, P_c(t) is the compressed pulse. L is the length of the whole Golay sequence/code. l is the width of a single pulse.

[0052] FIG. 7 shows the relationship between the object displacement and the S_ac at different noise levels, and the displacement measurement resolution of the system is almost unlimited if there is no noise. The white noise will only produce an offset to the relation between the displacement and S_ac. FIG. 7 shows the S_ac-displacement curves with different noise level in the echoes. The parameter settings of the simulation are the same as those in FIG. 3. The value of S_ac when the displacements equal to d=±7.5 μm, ±15 μm, . . . , ±75 μm are marked by the stars in each of the curves. The amplitude of the echoes is normalized to 1. In FIG. 7(a) there is no noise. In FIG. 7(b)˜(f) random noise is uniformly distributed in the interval (−0.2, 0.2), (−0.4, 0.4), (−0.6, 0.6), (−0.8, 0.8) and (−1, 1) respectively.

[0053] FIG. 8 shows the algorithm for the displacement measurement of a single target, including the following 9 steps: (1) Pulse compression of echoes A and B; (2) Pulse compression of echoes A and B with the shift ranging from −10 to 10 times of the sampling interval (step=1); (3) Calculating the amplitude of the sum of the absolute value of the clutters; (4) Find the minimum of the amplitude of the sum of the absolute value of the clutters (P0) (5) 10 times interpolation of pulse compression of echoes A and B; (6) Pulse compression of echoes A and B with offset from −1 (P1) to 1 (P2) times of the sampling interval (step=0.1); (7) Find the minimum of the amplitude of the sum of the absolute value of the clutters (P′0); (8) 100 times interpolation of pulse compression of echoes A and B; (9) Calculate the displacement.

[0054] In the measurement of the single target, the displacement was derived from the S_ac-displacement curve directly. The valid range of displacement measurement is determined by the monotonous zone of the S_ac-displacement curve (0˜75 μm in FIG. 7). The out-of-range displacement can be adjusted into the monotonous zone of the S_ac-displacement curve by adding a certain time shift to echoes of GCS B.

[0055] Based on the algorithm (FIG. 8), experimental data from an object is acquired by the system shown in FIG. 1 and analyzed. The actual displacements range from 1 nm to 6 μm in the experiment. The mean and STD of the error over all the measurements is −5.76 nm±36.27 nm. The actual displacement and the ultrasound measured displacement based on the clutter signal are compared in FIG. 9(a). The histogram of the measurement errors is shown in FIG. 9(b).

[0056] FIG. 10 shows the algorithm for displacement measurement in a desired region in an object. The algorithm is divided into three parts: mainlobe elimination, matched filtering, and amplitude normalization.

[0057] Mainlobe elimination: the purpose of mainlobe elimination is to extract the clutter signal from the compressed pulse. When the time interval between GCS A and GCS B is very short, the displacement in the desired region of the object is negligible. Therefore, the compressed pulse with very short time interval has only mainlobe and is subtracted from the compressed pulse with long time interval to eliminate the mainlobe.

[0058] De-convolution: the purpose of de-convolution is to accumulate the energy of the clutters and eliminate the overlap of the clutters. A pre-whitening filter is used to improve the SNR of the de-convolution. Pre-whitening aims to make the signal contains equal-strength components at all frequencies. A pre-whitening filter transform a non-white signal into a nearly white signal. This is performed by a predictor. The way the predictor whitens the signal is that it attempts to predict sample n based on the information from the previous samples. If we subtract this prediction from the actual sample n, we will be left with the portion of sample n that is not related to the rest of the samples.

[0059] The predictor is designed using AR model:

[0060] $\stackrel{⋒}{s}\left(n\right)=\sum _{k=1}^p{a}_{k}s\left(n-k\right)$

• • [0061] a_k is the coefficients of the AR model. a_k can be calculated according to the Yule-Walker equation:

[0062] $\left[\begin{array}{cccc}{R}_0& R_{-1}& \cdots & R_{1-p}\\ R_1& R_0& & R_{2-p}\\ ⋮& ⋮& & ⋮\\ R_{p-1}& R_{p-2}& \cdots & R_0\end{array}\right]\left[\begin{array}{c}{a}_1\\ a_2\\ ⋮\\ a_p\end{array}\right]=\left[\begin{array}{c}{R}_1\\ R_2\\ ⋮\\ R_p\end{array}\right]$

[0063] The transfer function of the pre-whitening filter is:

[0064] $H\left(z\right)=1-\sum _{k=1}^p{a}_k{z}^{-k}$

[0065] FIG. 11 show the coefficients of the pre-whitening filter for GCS (N=256). After the pre-whitening, a matching filter is then applied to the white signal to get output of the de-convolution.

[0066] Amplitude normalization: the purpose of the amplitude normalization is to normalize the amplitude of echoes from different objects/scatterers. A “scatterer” is an interface that scatters a transmitted wave, thereby returning at least some echo. A “target” refers to a selected area of study within an object, which area may contain one or more scatterers. The amplitude of clutters is proportional to both the amplitude of the echoes and the displacement. By amplitude normalization, the amplitude of clutters will only be proportional to the displacement.

[0067] The method for amplitude normalization is: (1) demodulate the output of matching filter (signal M) and the original compressed pulses with mainlobe (signal C) using Hilbert transformation; (2) divide the demodulated signal M by signal C.

[0068] Two computer simulations are done to verify the algorithm shown in FIG. 10.

[0069] In the first simulation, a 10 MHz, 90% bandwidth transducer is used. The transmission frequency is 10 MHz. The sampling rate is 100 MHz. The code length is 128. The sampling of echoes starts after the 128 code is transmitted. The echoes between the transmissions of the first code and the last code are not available. That produces a short undetectable zone called dead zone. The dead zone of this simulation is 9.6 mm. Ten scatterers are placed along the ultrasound beam (FIG. 12). The distances between the scatterers are: 1.5 mm, 2.25 mm, 3.0 mm, 3.75 mm, 4.5 mm, 5.25 mm, 6.0 mm, 6.75 mm and 7.5 mm, as shown in FIG. 12. From left to right, the displacement of the scatterers are 0.75 μm, 1.5 μm, 2.25 μm, 3 μm, 3.75 μm, 4.5 μm, 5.25 μm, 6 μm, 6.75 μm and 7.5 μm, respectively. The amplitude of the echoes from each of the scatterers are 1.1, 1, 0.9, 0.8, 0.7, 0.6, 0.5, 0.4, 0.3 and 0.2.

[0070] The results of the simulations of each step are shown in FIG. 13. FIG. 13(a) shows respectively: Echoes of GCS A, (b) Echoes of GCS B, (c) Compressed pulses, (d) Clutters signals derived after mainlobe elimination, (e) Results of De-convolution, (f) Results of amplitude normalization and demodulation.

[0071] FIG. 14 compared the amplitudes of the clutters (after the amplitude normalization) and the real displacement. FIG. 14 shows that the results of amplitude normalization is proportional to the real displacement.

[0072] In the second simulation, scatterers are distributed along the scan line evenly and the distance between scatterers is 225 μm. The amplitudes of the echoes from the scatterers have an exponential relation with the depth of the scatterers. The displacements of the scatterers have a sine distribution demonstrated in FIG. 15. The parameter settings of the ultrasound transducer and the GCS coded excitation are the same as those in the first simulation.

[0073] The results of the simulations of each step are shown in FIG. 16. FIG. 16(a) shows respectively: Echoes of GCS A, (b) Echoes of GCS B, (c) Compressed pulses, (d) Clutters signals derived after mainlobe elimination, (e) Results of De-convolution, (f) Results of amplitude normalization.

[0074] The absolute value of the real axis displacement settings and the envelope of waveform in FIG. 16(f) are compared in FIG. 17. The results show the algorithm described in FIG. 10 can measure very small displacement (±0.75 μm) in the desired reason of the object.

[0075] As various modifications could be made to the exemplary embodiments, as described above with reference to the corresponding illustrations, without departing from the scope of the invention, it is intended that all matter contained in the foregoing description and shown in the accompanying drawings shall be interpreted as illustrative rather than limiting. Thus, the breadth and scope of the present invention should not be limited by any of the above-described exemplary embodiments, but should be defined only in accordance with the following claims appended hereto and their equivalents.