Ultrasonic method and system for shear wave parameter estimation

[0002] The invention relates to a an ultrasonic method and an ultrasonic system for determining local propagation velocity of transient shear waves in a tissue, for displaying a sequence of velocity images of the transient shear waves and for determining tissue elasticity information.

[0003] The invention finds its application in using this information as a tool to diagnose abnormalities, such as tumors or edemas, in a patient tissue. These abnormalities are known to show changes of their mechanical properties with respect to sound background tissue. Shear wave propagation information permits of localizing said abnormalities.

[0004] A method for determining tissue elasticity in various parts of a body is already known from Sarvazyan, U.S. Pat. No. 5,606,971. According to this known method, ultrasonic waves are focussed at different location of a tissue, using a focused ultrasonic source that transmits said ultrasonic waves to its focal region. The focused ultrasonic source is preferably an ultrasound transducer of the phased array kind. Said focussed ultrasonic waves are amplitude modulated for generating shear waves at said different locations of the tissue. Said shear waves are further detected by measuring their amplitude and phase on the surface of the tissue. At least one propagation parameter of the shear waves in the tissue is determined from the phase and amplitude measures such as shear wave velocity, attenuation coefficient, amplitude and velocity of shear displacement of tissue particles in the propagating shear wave. A calculation, based on these measures, is performed and at least one mechanical parameter of tissue is determined such as the shear elasticity modules, Young modulus, dynamic shear viscosity, using known relations. The steps of the method are repeated for all amplitude modulated focused ultrasound waves, which are focused at said various locations. The calculated values of dynamic shear viscosity and elasticity modulus are displayed in function of the coordinates of said locations.

[0005] This known method of producing shear waves necessitates the use of focussed ultrasonic sources such as phased array transducers, for transmitting a great amount of ultrasonic energy to the locations where shear waves are produced in the tissue. Focussed ultrasonic sources may have destructive effects on the patient tissue due to the necessary amount of ultrasonic energy that is locally applied.

[0006] The present invention has for an object to provide a method, and a system to carry out the method, which do not use a focused ultrasonic source, in order to avoid possible secondary effects in the patient tissue. According to the invention, shear waves are generated in a tissue using an external mechanical vibration source. The shear waves produce displacements of tissue particles. According to the invention, a standard ultrasonic diagnostic system measures the velocity of the tissue particles and the velocity of the front of the shear waves. This ultrasonic diagnostic system also has means to display a sequence of images of the shear wave front propagation.

[0007] A problem is that the shear waves propagate at about 1 m. per s. over several centimeters in the tissue, for instance 4 cm. In this case, the propagation time is about 40 ms (milliseconds). This velocity is much too high and the propagation time delay much too small to permit of visualizing the effect of shear waves on the tissue using a standard sequence of ultrasonic images produced by a standard ultrasonic diagnostic imaging system. The image frame rate of such a system is not adapted to visualizing the shear wave propagation because it is of the order of 15 images per second whereas an image frame rate of about 1 image per ms (1000 images per second) is needed for shear wave visualization.

[0008] The present invention provides, as claimed in claim 1, an ultrasonic diagnostic imaging method for determining propagation parameters of transient shear wave front, comprising steps of forming transient shear waves in a tissue, acquiring ultrasonic image data (S,S*) of the tissue, along image lines, during a time delay for a transient shear wave front to propagate over a depth (z) in said tissue, estimating the tissue velocity (V) for each line, constructing a tissue velocity image sequence [I(V)] from the ultrasonic data (S,S*) and the tissue velocities (V) on the lines, and deriving the velocities (C_SW) of the shear wave front at instants of the sequence. The invention also provides an ultrasonic diagnostic imaging method, as claimed in claim 7, for determining tissue local mechanical parameters of a tissue from the transient shear wave front velocity (C_SW).

[0009] The present invention further provides a system, as claimed in claim 8 for carrying out said methods.

[0010] The invention allows the visualization of sequences of tissue moving under the influence of shear wave, using a standard ultrasound diagnostic imaging system with a standard transducer emitting and receiving standard ultrasound and echo signals respectively. The invention further allows the localization of tissue regions having contrasting mechanical properties with respect to a background, and the determination of the mechanical parameters of said tissue regions.

[0011] The invention is described in detail thereafter in connection with the schematic drawings in which:

[0012] FIG. 1A is a block diagram of an ultrasonic diagnostic imaging system having means for producing shear waves in a tissue, for processing echo signals and displaying corresponding images; FIG. 1B illustrates the scanning of ultrasonic lines in one image; and FIG. 1C illustrates the formation of an ultrasonic full-image sequence;

[0013] FIG. 2 is a block diagram of an ultrasonic diagnostic imaging system for processing ultrasonic signals of a tissue region under the influence of shear wave, for measuring the wave front velocity, and tissue mechanical local parameters;

[0014] FIGS. 3A and 3B illustrate respectively the formation of sequences of ultrasonic sub-images and the formation of corresponding sequences of tissue velocity sub-images;

[0015] FIG. 4 illustrates the construction of a sequence of tissue velocity full-images;

[0016] FIG. 5 illustrates an image constructed from the tissue velocity full-image sequence to provide the shear wave front velocity measure.

[0017] FIG. 1A shows a block diagram representing a standard ultrasonic diagnostic imaging system 1 connected to a transducer 10 denoted probe and to a vibration generator 2. The ultrasonic diagnostic system is connected to a processing system 100 for the determination of local propagation velocity of transient shear waves, for the display of a sequence of images the transient shear wave front and for the determination of tissue parameters such as tissue elasticity.

[0018] Referring to FIG. 1B, the probe 10 is positioned in ultrasonic contact with a tissue 5 to be examined and emits ultrasonic pulses through said tissue. The probe comprises transducer elements disposed in an arrangement parallel to the surface of the tissue and to an axis OX, denoted X-axis. The transducer elements permit of emitting several ultrasonic beams of ultrasonic pulses parallel to an axis OZ denoted Z-axis, orthogonal to the X-axis. The total number l_IMof ultrasonic beams may be for example l_IM=128 or l_IM=256. Each beam is hereafter referred to as an ultrasonic line parallel to the Z-axis and has an abscissa on the X-axis from x₁to xl_IM. As the ultrasonic pulses propagate through the tissue 5, along one line, at an abscissa x_iwith the index i such as 1≦i≦l_IM, corresponding echoes are generated by tissue particles encountered along said line. These echoes are received by the probe at instants that are function of the depth z of the echo formation on said line parallel to the Z-axis. The combination of all echoes generated from reflection along one line at an abscissa x_iforms a line of ultrasonic data of the tissue, which data are complex data and are denoted S and S*. The scanning of one line over a depth z of about 4 cm takes a time of less than or about Tl=100 μs; for the simplicity of explanation, it is supposed thereafter that this time is Tl=100 μs. The scanning of the l_IMlines is performed one line after the other and forms a full 2-D image of ultrasonic data (S and S*), referred to as ultrasonic image, which ultrasonic image comprises l_IMlines (128 or 256 lines) parallel to the Z-axis disposed at regular intervals (x₁, x₂, . . . xl_IM). The lines are about 4 cm long measured along the Z-axis. With the hypothesis set above, it takes about a time T_IM=l_IM100 μs to form one 2-D ultrasonic image over a depth of 4 cm:

[0019] T_IM=12,8 ms for one ultrasonic image of 128 lines,

[0020] or T_IM=25,6 ms for one ultrasonic image of 256 lines.

[0021] Referring to FIG. 1A, the ultrasonic diagnostic imaging system 1 is connected to a vibration generator 2 that is an external mechanical pulse generator 2. This external mechanical pulse generator presents the advantages of being very armless for the patient and very easy to use for the practitioner. Each mechanical pulse is applied to the tissue 5 by means of a contact body 4 and generates shear waves that propagates in tissue 5 over a depth z of about 4 cm at a velocity that has been A PRIORI estimated and that is:

[0022] C_SW=about 1 m/s

[0023] With this estimated velocity C_SW, the shear waves propagates in the tissue over said depth of 4 cm taking a time that is about:

[0024] T_SW=40 ms.

[0025] The invention proposes a method and a system for the visualization of the transient shear wave front on a display 45 connected to the standard ultrasonic diagnostic system 1. To that end, the ultrasonic diagnostic system of FIG. 1A comprises the processing system 100 associated to the standard ultrasonic diagnostic system 1, to the vibration generator 2 and to the probe 10. The probe 10 may be positioned with respect to the vibration generator 2 in any way appropriate for the standard ultrasonic diagnostic system 1 to form ultrasonic images of the tissue region 5 where the shear wave propagate.

[0026] For visualizing the transient shear wave front, first a temporal sequence of ultrasonic images is formed during the time delay of the shear wave front propagation. Referring to FIG. 1C, for visualizing correctly this propagation, it is proposed to form a temporal sequence of a number N of ultrasonic images, for instance N=10 to 50 images. In an example described thereafter, it is chosen to form a temporal sequence of:

[0027] N=40 ultrasonic images

[0028] denoted I₁, I₂to I_Nregularly acquired during the time delay:

[0029] T_SW=40 ms.

[0030] So, with the previous hypothesis related to scanning time Tl of one line, the ultrasonic diagnostic imaging system must supply one ultrasonic image every 1 ms to allow the visualization of the shear wave, whereas the standard ultrasonic diagnostic imaging system is able to supply only one ultrasonic image every 12,8 ms or every 25,6 ms, which is the time T_IM. So, such a standard ultrasonic diagnostic system is not adapted to supply an image frame rate appropriate to construct the temporal ultrasonic image sequence that is required to visualize the transient shear wave front propagation. The processing system 100 permits of solving this problem.

[0031] FIG. 2 shows a block diagram representing the system 100 having processing means for acquiring sequences of temporal ultrasonic sub-images, for estimating the local velocity of tissue particles under the action of the shear waves, for constructing a sequence of tissue velocity full-images, for estimating the shear wave front velocity and for visualizing the displacement of tissue under the influence of the transient shear wave propagation.

[0032] Referring to FIG. 3A, at a first instant α₁, the vibration generator generates a first shear wave in the tissue 5, and, using the scanning configuration 44 of the ultrasonic diagnostic system 1, at the same instant α₁, the probe 10 that is coupled to tissue 5 begins scanning a first line at the abscissa x₁in order to provide a first line of ultrasonic data denoted S and S*. It takes Tl=100 μs to the system to scan one line, and said system disposes of T_SW=40 ms to scan 4 cm, over which the shear wave propagates, and must supply N=40 temporal ultrasonic images at the rate of 1 ultrasonic image per ms. So, the system has time to scan a number 1<l_K≦10 adjacent lines of the first ultrasonic image I₁, disposed at abscissa x₁, x_{2, . . . x10}, thus forming a first band of lines of ultrasonic data (S and S*), which band is denoted K₁(I₁) in said first ultrasonic image I₁. For example, l_K=10 lines.

[0033] Referring to FIG. 3A, the system further scans a number l_K=10 adjacent lines of the second temporal ultrasonic image I₂, disposed at the same abscissa x₁, x₂, . . . x₁₀, thus forming a first band of lines of ultrasonic data (S and S*), denoted K₁(I₂) of l_K=10 lines at the same abscissa in the second ultrasonic image I₂.

[0034] Referring to FIG. 3A, the system further scans l_K=10 adjacent lines of each temporal ultrasonic image including I₄₀, forming in each temporal ultrasonic image I₁to I₄₀only one band K₁of l_K=10 lines disposed at the abscissa x₁to x₁₀. The first band K₁is denoted first sub-image K₁of each completed ultrasonic image of 128 or 256 lines to be performed. It takes T_SW=40 ms for the system to perform N=40 temporal ultrasonic sub-images K₁(I₁) to K₁(I₄₀) of ultrasonic data S and S*. This forms a first sequence of ultrasonic sub-images of the ultrasonic data (S and S*).

[0035] As it may be understood of those skilled in the art, the number of lines of the ultrasonic sub-image is calculated: by determining the total number l_SWof lines that are possibly scanned during the time propagation T_SWof the shear wave, knowing the time Tl for scanning one line, by fixing the number N of ultrasonic images of the sequence that is appropriate to visualize the shear wave front displacement, which gives the number l_Kof lines possibly scanned per ultrasonic image, from I₁to I_N, during the time of one pulse of the vibration generator 2:

[0036] l_K=l_SW/N

[0037] At a second instant α₂, the vibration generator 2 generates a second shear wave in the same region of tissue 5, and, using the scanning configuration 44, at the same instant α₂, the probe 10, that is coupled to tissue 5, begins scanning a first line at the abscissa x₁₁, which forms the first line of ultrasonic data (S and S*) of a second band K₂(I₁) of the first temporal ultrasonic image I₁. So, now, the system scans l_K=10 lines at the abscissa x₁₁to x₂₀, thus forming the second band of ultrasonic data (S and S*), which band is denoted K₂(I₁) and is composed of l_K=10 lines in the first temporal ultrasonic image I₁.

[0038] Then as previously described in reference to FIG. 3A, the system further scans l_K=10 adjacent lines of each temporal ultrasonic image I₁to I₄₀, disposed at the abscissa x₁₁to x₂₀, forming the bands of temporal ultrasonic sub-images K₂(I₁) to K₂(I₄₀).

[0039] At further instants α₃to α_k, the vibration generator 2 provides a mechanical pulse that generates a shear wave propagating in the same region of tissue 5 and, using the scanning configuration 44, at the same instant, the probe 10 that is coupled to the tissue 5 begins scanning a first line of l_K=10 lines to form the ultrasonic image of a corresponding band K of the temporal ultrasonic images I₁to I₄₀.

[0040] This scanning operation is performed until a number k of bands or ultrasonic sub-images K₁to K_kare constructed for each ultrasonic image I₁to I₄₀. The number k of ultrasonic sub-images and the number l_Kof lines in each ultrasonic sub-images are chosen appropriately by the user in function of the number of lines l_IMn of each full image. At the instant α_k, the last sequence of ultrasonic sub-images is constructed.

[0041] Referring to FIG. 2, the ultrasonic data of the sequences of ultrasonic sub-images are used to construct corresponding sequences of tissue velocity sub-images. The velocities of tissue particles under the action of shear waves are measured along each scanning line of the ultrasonic sub-image sequence. The measure of velocity of displacement of tissue is an operation well known of those skilled in the art, which is usually performed using the standard ultrasonic system 1 that has means 41 for tissue velocity measurement.

[0042] The ultrasonic diagnostic imaging system 1 may perform the tissue velocity measurement on each scanning line at the abscissae x₁to xl of the ultrasonic sub-images of the sequences respectively K₁to K_kusing different methods. A known method is for example the “phase shift method” according to which the tissue velocity [V(n,l,z)] is supplied by the processing means 41, from the ultrasonic data S and S* of the lines of the sequences of ultrasonic sub-images, according to the formula: $V\left(n,l,z\right)=\frac{C}{4\pi \left(l_K{T}_l\mathrm{fo}\right)}\mathrm{Arg}\sum _{m=n-p}^{m=n+p}S\left(m,l,z\right)S*\left(m+1,l,z\right)$

[0043] where n is the image number in the sequence of N ultrasonic sub-images, with 1≦n≦N−1, where l is the line number in the considered ultrasonic sub-image among the l_Ktotal number of lines in said ultrasonic sub-image K₁to K_kof a given ultrasonic sub-image n in which the velocity of tissue is estimated, Tl is the scanning time of one line, fo is the mean frequency of the echo signal, S and S* are the values of the complex ultrasonic signals, z is the depth of the point where the velocity is estimated and C is the velocity of ultrasounds in the tissue (in the present example, 1≦n≦39, l_K=10 and Tl=100 μs). In the above formula, p is the number of data that are averaged in order to calculate the velocity; p may be taken equal to 2 for example; m is calculated from n and p.

[0044] Referring to FIG. 3B, with this method, sequences of tissue velocity sub-images are constructed, that are composed of a number N−2p (for example N−2p=40−4=36) of sub-images, which is different from the number N of sub-images (N=40) of the sequences of ultrasonic sub-images. However the number k of sequences of tissue velocity sub-images is the same as the number of sequences of ultrasonic sub-images.

[0045] In a variant, the tissue velocity may be measured by a well known “temporal shift method”, or other methods that are the choice of the user.

[0046] Then, the sequences of tissue velocity sub-images are stored in a first memory means 42, denoted MEM1, and notably the tissue velocity values are stored as a function of their locations, denoted P(x,z,t) where x is the abscissa of the scanning line along the X-axis, t is the instant, which is different from n, corresponding to said image number n and z is the depth along the Z-axis. So P(x,z,t) is a point in a tissue velocity sub-image at an instant t corresponding to n at the abscissa x and depth z, and in a sub-image K₁[V(n,l,z)] to K_k[V(n,l,z)] of the k sequences of (N−2p) tissue velocity sub-images.

[0047] Referring to FIG. 2, the k tissue velocity sub-images K₁[V(n,l,z)] to K_k([V(n,l,z)] are further processed in order to be assembled for constructing a temporal sequence of N−2p tissue velocity full-images, denoted I₁[V(n,l,z)] to I_N−2p[V(n,z)].

[0048] This operation is performed using a first processing means 40 denoted PROCESSING1 of the processing system 100. The PROCESSING1 operates a synchronization of the lines of all tissue velocity sub-images for forming tissue velocity full-images. In this operation, all the first lines of all the k tissue velocity first sub-images, which are adjacent, are synchronized to the first line of the first sub-image formed at the first instant α₁, in order to cover a first full-image, and all the lines of all the k tissue velocity sub-images corresponding to said given full-image are synchronized on said first line, taking into account the time delay between the formation of each image line. Then, all the adjacent sub-images of the tissue velocity sequences are processed in the same way with respect to the instant of formation of the first sub-images among the adjacent sub-images, for covering tissue velocity full-images.

[0049] Referring to FIG. 4, the resulting tissue velocity sequence is then constituted by the number N−2p=36 tissue velocity full-images I₁[V(n,l,z)] to I_N−2p[V(n,l,z)] that are so constructed. By this operation of synchronization, each tissue velocity full-image I₁[V(n,l)] to I_N−2p[V(n,l)] of the tissue velocity sequence corresponds to a respective instant t of shear wave propagation in the tissue, t being function of n previously defined.

[0050] In order to perform the synchronization operation, the velocity values related to the lines, which are stored in MEM1 are extracted and input to the synchronization means PROCESSING1, which has for a function to correct the time delay produced by the scanning operation. First, in order to construct one full-image at one instant t of the tissue velocity sequence, the lines of each tissue velocity sub-image of the k tissue velocity sequences must be synchronized to the first line of the first tissue velocity sub-image corresponding to said tissue velocity full-image. Second, the point on the lines of said full-image must be synchronized to corresponding tissue velocity values estimated from the sequences of tissue velocity sub-images, taking into account that the sequences of ultrasonic sub-images comprise N ultrasonic sub-images presented at instants n whereas the sequences of tissue velocity sub-images comprise N−2p tissue velocity sub-images presented at instants t. It is to be noted that the real time delay between instants tl and tl₊₁corresponding to the velocities on two adjacent lines at the same depth z, respectively V(n,l,z) and V(n,l+1,z), is Tl, which is the time delay for scanning one line. It is also to be noted that the time necessary to scan one ultrasonic sub-image is l_KTl. So, the time delay between the instants t1 and tl corresponding to the velocities V(n,1,z) and V(n,l,z) is:

[0051] tl−t1=(l−1)Tl.

[0052] For the above-described reasons, finding the actual velocity on a given line numbered l in the tissue velocity sub-image is performed by using an interpolation function, denoted I, of the velocity data in function of time t, according to the formula:

[0053] fl(t)=I[V(n,l,z)] from n=p to N−2p

[0054] where

[0055] t=[n+(l−1)]l that is smaller than l_KTl.

[0056] This function fl (t) is a filtering function applied on the velocity data that provides the actual tissue velocity on a considered line of the tissue velocity image to be constructed. This filtering function may be of a kind that is known by those skilled in the art, for example a cubic spline function or a sinus cardinal function.

[0057] The number k of sub-images K, the number N of temporal ultrasonic images and the number l of lines to form a temporal ultrasonic image are the choice of the user. At the end of this phase 3, a temporal sequence of a number N−2p=40−2p of tissue velocity full-images has been constructed from the k acquired sequences of N ultrasonic sub-images and the estimation of the tissue velocity.

[0058] Referring to FIG. 2, the data issued of PROCESSING1 are stored in a second memory means 47 denoted MEM2. The full-images are tissue velocity images of the region 5 where the shear waves propagate, and are displayed in sequence using the display means 45 of system 1. At this stage of the process, the display shows the tissue regions moving under the action of the mechanical pulse supplied by the vibration generator 2.

[0059] Referring to FIG. 2, in a further phase, the data of the tissue velocity image sequence are extracted from MEM2 and supplied to a second processing means 50, denoted PROCESSING2. From the data of the tissue velocity sequence, which comprise the tissue velocity values at each point of each image, and at each instant t of the shear wave propagation, one further image, referred to as I_MAX, is constructed in PROCESSING2.

[0060] Referring to FIG. 5, in PROCESSING2, this new image I_MAX, called image of the maximum velocities, is build by writing, at points on lines of said image, the instants, referred to as t_MAX, among t₁to t_N−2p, when the velocities of the tissue are maximum, denoted V_MAX. The different points of the image I_MAXthat correspond to maximum of tissue velocities at the same instant t_MAXare joined by lines denoted L_t: all the points on a same line are points of maximum velocities for a same instant t: the lines are equal-levels of velocities. So, these lines show the instants of propagation of the shear wave front. The estimation of the gradients of time in this image I_MAXpermits of determining the velocity of said shear wave front. The wave front velocity, denoted C_SWis estimated in PROCESSING2, in function of the gradients in said new image I_MAX, by the following formula: $C_{\mathrm{SW}}={\left(\frac{d{t}_{\mathrm{MAX}}^2}{dx}+\frac{d{t}_{\mathrm{MAX}}^2}{dz}\right)}^{-\frac{1}{2}}$

[0061] The propagation front velocity data extracted from PROCESSING2 are further processed in a calculation means 51 for estimating the tissue parameters such as for example the elasticity of tissue regions from formulae known of the document cited in the introductory part. The tissue parameters are stored in memory means 52 in order to be further used as diagnostic tools for helping to the diagnostic of tumors or other diseases.

[0062] This ultrasonic diagnostic imaging system may be a standard ultrasonic apparatus 1,10 associated to a processing means 100 comprising a suitably programmed computer, a processor of a workstation, or a special purpose processor having circuit means such as LUTs, Memories, Filters, Logic Operators, that are arranged to perform the functions of the method steps according to the invention. The workstation may also comprise a keyboard, a screen 45 and a mouse. The processing system may be connected to supplementary storing means to store medical images.