Patent Grant
9952321
This is the U.S. National Phase of International Application No. PCT/FR2006/000702filed 30 Mar. 2006, the entire disclosure of which is incorporated herein by reference.
The present invention relates to imaging methods and devices using shear waves.
More particularly, the invention relates to an imaging method in which a mechanical wave having a shear component and a compression component is generated in a viscoelastic medium and at least one parameter describing the movement of the viscoelastic medium at various points during the propagation of this mechanical wave is determined.
Thus, a qualitative and/or quantitative analysis may be performed, in particular to identify regions of different hardness from the rest of the viscoelastic medium or regions having a different relaxation time from the rest of the viscoelastic medium.
Document WO-A-04/21038 describes an example of such a method, in which said parameter describing the movement (especially the displacement) in a plane, following the remote generation of a localized pulsed thrust, which creates the mechanical wave in question, in this case a shear wave, is determined.
Although this method is already completely satisfactory, the present invention disclosure will further perfect it, so as to improve its precision and, where appropriate, to use it for imaging in three dimensions.
For this purpose, a method of the kind in question is characterized in that it includes a correction step during which the movement parameter is processed so as to remove the effects due to the compression component of said mechanical wave.
Thus, even when a pulsed thrust is used, if the compression component of the elastic mechanical wave that results therefrom disturbs the measurement of the movement parameter, only the effects of the shear component are taken into account. When instead of a pulsed mechanical wave a sustained wave is used, for example in order to carry out a relatively long measurement, such as for three-dimensional imaging, the compression component of the mechanical wave is still present during the time of measuring the movement parameter and it therefore greatly disturbs this measurement. The processing according to the invention makes it possible to eliminate this disturbance and, where appropriate, to map the viscoelastic parameters of the medium in two or three dimensions.
In preferred embodiments of the method according to the invention, one or more of the following arrangements may optionally be furthermore employed:
Moreover, the subject of the invention is also an imaging device comprising means for generating, in a viscoelastic medium, a mechanical wave having a shear component and a compression component and means for determining at least one parameter describing the movement of the viscoelastic medium at different points during the propagation of this mechanical wave, characterized in that it includes correction means for processing the movement parameter so as to remove the effects due to the compression component of said mechanical wave
In preferred embodiments of the device according to the invention, one or more of the following arrangements may furthermore be optionally employed:
Other features and advantages of the invention will become apparent over the course of the following description of one of its embodiments, given by way of non-limiting example, in conjunction with the appended drawings.
In the drawings:
In the various figures, identical references denote identical or similar elements.
The imaging device 1 shown in
In the example shown in the drawings, the imaging device 1 is more especially designed for medical imaging of the breast of a female patient 3, the breast therefore constituting the medium 2 to be imaged.
For this purpose, the imaging device 1 may for example include an examination table 4 on which the patient may lie on her stomach. This table is provided with an opening 5 allowing at least one breast 2 of the patient to pass through it, which breast hangs in a bath 6 of ultrasonic coupling agent, for example a gel or liquid, preferably water This liquid bath 6 is bounded by a vessel 7 that also contains an array 9 of ultrasonic transducers and a vibration generator 8 in contact with the breast. The vessel 7 may be transparent, in particular to check that the vibration generator 8 is properly positioned against the breast 2 and that the transducer array 9 is properly positioned alongside the breast.
It should be noted that the examination table 4 may, where appropriate, be designed to allow both the patient's breasts to pass through it, these then hanging in the same liquid bath 6 or in two different baths Therefore the imaging device may be optionally adapted for imaging both breasts of the patient simultaneously.
The vibration generator 8 may comprise an acoustic transducer 10 which generates low-frequency mechanical vibrations in a height-adjustable plate 11, for example positioned in contact with the end of the patient's breast 2 in order to transmit mechanical (elastic) waves thereinto, especially sustained waves comprising a compression component and a shear component. The mechanical waves in question may have for example a frequency between 0.1 Hz and 10 kHz.
It should be noted that the abovementioned mechanical waves could be generated other than by the abovementioned vibration generator, fox example:
In the example shown in
As shown in
The transducer array 9 may for example be mounted on a rotary platform 15 supported by a fixed platform 16. The two platforms 15, 16 lie in the horizontal plane (X-Y) and are advantageously apertured in order to allow the patient's breast 2 to pass freely through a central passage 17. The rotary platform 15 can be moved angularly with respect to the fixed platform 16, about a vertical central axis Z0, by means of an adjustment system operated by a sealed electric stepper motor (not visible in
The transducer array 9 may furthermore be able to be moved translationally in the directions indicted by the double arrow 19, perpendicular to the analysis plane (D-Z), in order to image several parallel analysis planes (for example about thirty analysis planes) before said analysis plane is rotated. For this purpose, the array of transducers may for example be mounted so as to slide along the directions 19 on a transverse beam 20 that extends along a direction orthoradially to the central axis Z0 and is fixed to the rotary platform 15. The transducer array 9 moves along the beam 20, for example by means of an adjustment system operated by a sealed stepper motor controlling a nut/screw assembly incorporated into the beam 20.
The transducer array 9 may, where appropriate, be connected to the beam 20 via a swivel 21 which slides along said beam and on which the array 9 can pivot about a rotation axis R parallel to the abovementioned direction 19. It is thus possible, optionally, to vary the direction D in elevation, that is to say to vary the angle of inclination of the direction D to the horizontal plane (X-Y) without varying the analysis plane (D-Z). As will be explained below, this optional movement may also vary the points of view along which the breast 2 is imaged, so as to reconstruct an image thereof in three dimensions.
The pivoting of the transducer array 9 may for example be controlled by a sealed electric stepper motor 22, allowing the array 9 to be displaced in the direction defined by the double arrow 23.
Optionally, the fixed platform 16 may also be able to be moved vertically, in order to image the breast at several heights, especially if the array 9 of transducers is not long enough to cover the desired height of the observation field in the breast 2.
In the example shown, the height of the fixed platform 16 may for example be adjusted by means of a screw cylinder 24 operated by a sealed electric stepper motor or the like, connected to the platform 16, for example by a support rod 25, in order to move the platform 16 in the direction defined by the double arrow 26, along the vertical axis Z.
As shown in
a microprocessor or the like—of an electronic signal processing rack 35.
The central processing unit 34 also controls the n transducers 36 (T1, T2, . . . Tn) of the transducer array 9 (n may for example be 128 or more).
Optionally, the rack 35 may also include:
The device that has just been described operates as follows.
The microcomputer 27 generates sustained elastic mechanical waves in the viscoelastic medium 2 by the acoustic transducer 10 during an excitation step (a).
Moreover, the propagation of the mechanical wave in the viscoelastic medium 2 is observed by echography during an observation step (b) concomitant with the excitation step (a) For this purpose, the microcomputer 27 also causes ultrasonic compression waves to be transmitted (at a frequency for example between 0.5 and 100 MHz, and preferably between 0.5 and 15 MHz, for example around 4 MHz) via the transducers 36 of the array 9.
These compression waves, after they have passed through the coupling gel or liquid 6, penetrate into the medium 2 where they are reflected off scattering particles contained in said medium 2 thereby allowing the movements of the medium 2 to be monitored. The scattering particles in question may consist of any heterogeneity in the medium 2 and especially, when medical application is involved, by collagen particles present in human tissue.
The compression waves thus transmitted may for example be “plane” ultrasonic compression waves (i.e. in this case a wave in which the wavefront is a straight line in the analysis plane (D-Z)) or any other type of focused or unfocused wave clarifying the entire field of observation in the medium 2, for example a wave generated by transmitting random acoustic signals via various transducers 36.
The abovementioned observation step (b) may comprise the following substeps, for each position of the transducer array 9:
During step (b1), the ultrasonic compression waves are transmitted at a rate of between 10 and 10 000 shots per second and preferably between 10 and 5000 shots per second (this rate is limited by the round-trip time of the compression wave in the medium 2, and therefore by the thickness of the medium 2 in the direction D. This is because it is necessary for all the echoes generated by the compression wave to be received by the probe 36 before a new compression wave is sent).
Each ultrasonic compression wave propagates in the medium 2 with a much higher propagation velocity than the shear waves (for example around 1500 m/s in the human body) and interacts with the scattering particles of the medium 2, thereby generating echoes or other analogue perturbations of the signal, known per se by the term “speckle noise” in the field of echography.
This speckle noise is picked up by the transducers T1 . . . Tn during substep (b2), after each ultrasonic compression wave firing. The signal sij(t) thus picked up by each transducer Ti after the jth firing is firstly sampled at high frequency (for example a frequency ranging from 30 to 100 MHz) and digitized in real time (for example over 12 bits) by the corresponding sampler Ei.
The signal sij(t) thus sampled and digitized is then stored, also in real time, in the corresponding memory Mi.
Again for each position of the transducer array 9, the microcomputer 27 then carries out, generally at a later time, a processing step (c) during which:
During substep (b2), after all the signals sij(t) corresponding to a measurement in the same position of the transducer array 9 have been stored, the central processing unit 34 causes these signals to be reprocessed by the summer circuit S (or else said central processing unit itself carries out this processing, or said processing may be carried out in the microcomputer 27) by a conventional beamforming procedure corresponding to substep (c1).
Thus, signals Sj(r,z) each corresponding to the image of the field of observation in the plane (D-Z) after the jth firing are generated.
For example, a signal Sj(t) may be determined by the following formula:
where:
After the optional beamforming step, the central processing unit 34 stores the image signals Sj(r,z), where j is the number of the compression wave firing, in a central memory M belonging to the rack 7. These signals may also be stored in the microcomputer 4 when the latter itself carries out the image processing.
These images are processed at a later time or immediately after substep (c2), by correlation and advantageously by intercorrelation, either pairwise or preferably with a reference image which may be:
The abovementioned intercorrelation may for example be carried out in the circuit DSP forming part of the Tack 35 or be programmed in the central processing unit 34 or in the microcomputer 27.
During this intercorrelation process, an intercorrelation function <Sj(x,y), Sj+1(x,y)> for example is maximized so as to determine the displacement undergone by each particle 5 giving rise to an ultrasonic echo.
Examples of such intercorrelation calculations are given in the prior art, especially by O'Donnell et al. (“Internal displacement and strain imaging using speckle tracking”, IEEE Transactions on Ultrasonics, Ferroelectrics and Frequency Control, Vol. 41, No. 3, May 1994, pp 314-325) and by Ophir et al. (“Elastography: a quantitative method for imaging the elasticity of biological tissues”, Ultrasonic imaging, Vol. 13, pp 111-134, 1991).
For generalizing the intercorrelation technique using the method proposed by Tanter et al. (M. Tanter, J. Bercoff, L. Sandrin and M. Fink, “Ultrafast compound imaging for 2D motion vector estimation: Application to transient elastography”, IEEE, Transactions Ultrasonics, Ferroelectrics and Frequency Control 49 (10), pp 1363-1374, 2002), for each position of the transducer array 9, a set of displacement vectors {right arrow over (u)}({right arrow over (r)},t) generated by the shear waves at each position {right arrow over (r)} in the field of observation in the analysis plane (D-Z) under the effect of the shear wave is thus obtained (these displacement vectors are in two dimensions and have a component along the direction D and a component along the direction Z).
It is also possible to carry out the measurement or to refine the above technique for measuring the displacements (or to enlarge the field of observation) in the analysis plane (D-Z) by carrying out another series of ultrasonic wave shots while modifying the site of the transducer array 9 by pivoting said array about the abovementioned axis R.
Measurement of the displacements is then repeated in several parallel analysis planes (D-Z) by displacing the transducer array 9 into several successive positions along the beam 20 It is thus also possible to measure the displacement, for example in around thirty parallel planes in the case of breast imaging, in a total time of about 1 to 2 s for all the parallel planes.
The platform 15 is then rotated though a predetermined angle and thereby also rotating the analysis plane (D-Z), and the abovementioned operations of measuring displacements in several analysis planes parallel to the new direction D as explained above, are repeated. Displacement fields {right arrow over (u)}({right arrow over (r)},t) in two dimensions, with a component along the new direction D and a component along the direction Z, are thus obtained.
After having measured the displacements with at least two orientations of the direction D (for example two orientations at 90° to each other) or where appropriate more than two orientations (for example up to thirty-six angular orientations), a field of displacements {right arrow over (u)}({right arrow over (r)},t) in three dimensions in the viscoelastic medium 2 may be reconstructed.
To do this, the values of the displacements, for example along each axis X, Y, Z, as a function of time are collected, where appropriate after temporal resetting of the various measurements (for this purpose, a signal may be transmitted by the transducer 10 with a periodic modulation and the time origin may be taken as one and the same point in the periodic modulation before each sequence of ultrasonic wave shots), and then these values are combined by averages and interpolations to a single field of displacement vectors {right arrow over (u)}({right arrow over (r)},t) in three dimensions.
Next, this field of displacements {right arrow over (u)}({right arrow over (r)},t) (or another movement parameter) is corrected in order to eliminate therefrom the effects due to the compression component of the elastic mechanical wave generated by the acoustic transducer 10. For this purpose, the curl (rotational) operator is applied to the field {right arrow over (u)}({right arrow over (r)},t) in order to determine a vector field {right arrow over (q)}({right arrow over (r)},t)=
Thus, only the effects of the shear component of the elastic mechanical wave generated by the acoustic transducer 10 are taken into account.
This set of vectors {right arrow over (q)}({right arrow over (r)},t) is stored in the memory M or in the microcomputer 4 and may for example be displayed, especially by means of the screen 4a of the microcomputer, in the form of a slowed-down film where the value of the components of q is illustrated by an optical parameter, such as a gray level or a chromatic level.
Thus, the differences in propagation of the shear wave between the regions of different characteristics in the medium 2, for example the healthy tissue and cancerous tissue in the case of a medical application, may be perfectly displayed.
This shear wave propagation film is furthermore superposable with a conventional echographic image, which may also be generated by the device 1 described above.
Moreover, it is also possible to calculate, not the displacements of each point in the observed medium 2, but the strains of the medium 2, that is to say vectors whose components are the derivatives of the components of the respective displacement vectors with respect to the space variables. These strain vectors can be used, like the displacement vectors, for clearly displaying the propagation of the shear wave in the form of a film and they also have the advantage of factoring out the displacements of the probe 36 relative to the observed medium 2. These strain vectors are processed by applying the curl (rotational) operator to eliminate the effects due to compression component of the elastic mechanical wave generated by the acoustic transducer 10, as explained above in the case of the displacement vectors.
Next, the microcomputer 27 may advantageously carry out a mapping step (d) based on the displacement field or strain field, during which step at least one parameter describing the propagation of the shear waves in the medium 2 is calculated from the variation in the movement parameter (displacement, displacement or strain velocity) processed by applying the curl operator.
The shear wave propagation parameter calculated during the mapping step is for example chosen from: the velocity C1 of the shear waves, the shear modulus μ or the Young's modulus E(=3μ), the attenuation of the shear waves, the shear elasticity, the shear viscosity ζ, or the mechanical relaxation time of the tissue, or else several of these parameters in the case of an an isotropic medium.
For example, it is possible to calculate the shear modulus μ and the shear viscosity ζ at various points in the field of observation.
To do this, the following propagation equation (1) may be used (assuming that the medium 2 may be considered to a first approximation as a locally homogeneous and isotropic viscoelastic medium), which the displacements {right arrow over (u)}({right arrow over (r)},t) generated by the elastic mechanical waves at each position {right arrow over (r)} in the medium 2 obey:
ρ∂12{right arrow over (u)}=μ∇2{right arrow over (u)}+(λ+μ)∇(∇{right arrow over (u)})+ç∂1∇2{right arrow over (u)}+(ξ+ç)∂1∇(∇{right arrow over (u)}), (1)
where ρ is the density of the medium, μ is the shear modulus (that we seek to reconstruct), λ is the second lamé coefficient, ç is the shear viscosity, taking into account the dissipation into the medium, and ζ is the viscosity of the compression wave. The shear modulus μ and the second Lamé coefficient λ are given, in terms of Young's modulus E and Poisson's ratio σ by the equations:
Each vector field ({right arrow over (u)}) may be decomposed into an irrotational part ({right arrow over (u)}I), an incompressible part ({right arrow over (u)}T) and a harmonic part ({right arrow over (u)}H) which is both irrotational and incompressible. The harmonic part {right arrow over (u)}H takes into account the global displacements and can therefore be neglected, since here we are studying the case of waves. Using this decomposition, it is possible to show that equation (1) splits into two equations expressing the propagation of the shear waves and of the compression waves with their respective propagation velocities cT and cL (ignoring the effect of the viscosity), i.e:
By applying the curl operator to equation (1), a simplified wave equation is thus obtained that no longer contain a contribution from the compression component of the elastic mechanical wave generated by the transducer 10, namely:
ρ∂12{right arrow over (q)}=μ∇2{right arrow over (q)}+∂1ç∇2{right arrow over (q)}, {right arrow over (q)}=∇×{right arrow over (u)}, {right arrow over (q)}∈C3 (4)
It is then possible to calculate, from equation (4), the parameters μ and ζ and the price, now, of third-order spatial derivatives. Equation (4) can be solved simply, for example by being multiplied by the transpose matrix of the system (pseudo-inverse). It is noted that equation (5) represents in fact six equations (three for the real part and three for the imaginary part of {right arrow over (q)}, in complex notation) for only two unknowns (the density ρ is considered to be constant and equal to that of water).
Thus, in the case of the steady state, with the following notations:
qiR=Ai cos(ωt+φi)|t=0, qiI=Ai cos(ωt++φi)|t=π/(2ω), i=x,y,z (5)
the available equations may be expressed as:
Multiplication by the transpose matrix gives:
with i=x,y,z and the Einstein convention for the identical indices. In the case of the steady state, there is no correlation between μ and ζ, thereby making it easier to map these two calculated parameters, which are therefore independent of each other.
It should be noted that similar calculations may be used to determine the desired propagation parameters even when the approximation of a locally homogeneous and isotropic medium cannot be made. In this case, a more complex propagation equation is used, namely:
where μi is the ith component of the displacement vector {right arrow over (u)}({right arrow over (r)},t), ρ is the density of the medium 2, λiklm is the elasticity tensor of rank 4, and ηiklm is the viscosity tensor of rank 4 (using the Einstein convention for the identical indices)
This equation (1′) simplifies where the assumption of a transversely isotropic medium, that is to say a medium having fibers aligned in only one direction. In this case the parameters describing the material may be divided into two groups:
The relationship between the strain tensor and the stress tensor is given here by the equation:
where uik=½(∂xkui+∂xiuk) is the strain tensor and σik is the stress tensor.
The corresponding propagation equation for the displacement vector u is given by:
where ζ is the shear viscosity. The viscosity of the longitudinal waves (ξ) may be neglected at relatively low frequencies (in the Hz or kHz range) In addition, ∇u has a low value owing to the value of the Poisson's ratio. Thus ξ, ζ and ∇u have low values and the third part on the right-hand side of equation 9 can therefore be legitimately neglected Putting equation 8 into equation 9 gives:
with the notation τ=−μ⊥ and assuming that all the elastic moduli of the longitudinal waves are equal, i.e. λ=
=λ⊥=λM.
After application of the curl operator, the following equation is obtained:
This equation (11) gives in fact six equations that can be used in particular to calculate μ⊥, τ and ζ.
It should be noted that the invention is applicable whatever the method of measuring the movement parameter (displacement or the like) of the medium 2. This parameter would thus be, where appropriate, determined by MRI.
The method that has just been described could also be coupled with acoustooptic imaging of the medium 2, as described for example in document WO-A-04/85978.
| Number | Date | Country | Kind |
|---|---|---|---|
| 05 03376 | Apr 2005 | FR | national |
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/FR2006/000702 | 3/30/2006 | WO | 00 | 10/2/2007 |
| Publishing Document | Publishing Date | Country | Kind |
|---|---|---|---|
| WO2006/106213 | 10/12/2006 | WO | A |
| Number | Name | Date | Kind |
|---|---|---|---|
| 5606971 | Sarvazyan | Mar 1997 | A |
| 5810731 | Sarvazyan et al. | Sep 1998 | A |
| 5855557 | Lazenby | Jan 1999 | A |
| 6791901 | Robertsson et al. | Sep 2004 | B1 |
| 20040204860 | Ghosh et al. | Oct 2004 | A1 |
| 20050038339 | Chauhan et al. | Feb 2005 | A1 |
| 20050252295 | Fink et al. | Nov 2005 | A1 |
| Number | Date | Country |
|---|---|---|
| WO 2004021038 | Mar 2004 | WO |
| Entry |
|---|
| French Search Report FR 0503376 dated Nov. 16, 2005. |
| International Search Report for parent patent app. PCT/FR2006/000702 dated Jun. 22, 2006. |
| Tanter—Ultrafast Compound Imaging for 2-D Motion Vector Estimation: Application to Transient Elastography, IEEE Transactions on Ultrasonics, Perroelectrics and Frequency Control, vol. 49, No. 10, Oct. 2002. |
| Number | Date | Country | |
|---|---|---|---|
| 20090124901 A1 | May 2009 | US |
