Method and system for ultrasound characterization of media for medical analysis

Abstract

The invention relates to a method for the ultrasonic characterization of a medium, comprising the steps of generating a series of incident ultrasonic waves, measuring a normative reflection matrix Rui (t) defined between a transmission basis (i) at the input and a reception basis (u) at the output, determining a focused reflection matrix of the medium by focusing the normative reflection matrix, with virtual transducers located at the same depth. The method then comprises filtering the focused reflection matrix Rxx (z) to obtain a single or multiple reflection matrix, and then determining a single or multiple scattering rate local to the spatial location point rp. The single or multiple scatterers are local physical characteristics representative of local properties of the medium.

Description

Ultrasonic characterization method and system for media used in medical analysis Technical Field

[0001] This disclosure relates to methods and systems for ultrasound characterization of media, intended for use in medical imaging. These methods and systems implement a transducer array placed in contact with, for example, a living organism, animal, or patient, to emit ultrasound waves into the body and measure the reflection from the scattering medium caused by the emission. Background Technology

[0002] In medical ultrasound imaging, certain physical parameters characterizing wave propagation in a medium are measured to detect or monitor the progression of pathology. Attenuation is particularly important; its characteristic length is denoted as l. ext The extinction length corresponds to the characteristic distance over which the wave amplitude decays as it propagates through the medium, and depends in particular on two factors: absorption and scattering.

[0003] Wave scattering is a phenomenon used in ultrasound imaging to probe the medium under study. Each scattering event, caused by inhomogeneities within the medium, produces an echo, which, under certain conditions, can be used to construct a reflectivity image of the medium. Each scattering event also causes attenuation of the incident wave, with a characteristic distance of scattering mean free path ls, corresponding to the average distance between two consecutive scattering events.

[0004] Absorption also causes attenuation of the incident wave, with a characteristic distance being the absorption length l. a .

[0005] These two phenomena lead to the overall attenuation of the wave within the medium, making Attenuation varies depending on the nature of the tissue being examined and produces observable contrast differences in conventional ultrasound images. These contrast differences can be corrected globally and approximately by applying depth gain compensation to the reconstructed ultrasound images.

[0006] In medical imaging practice, attenuation throughout the entire acoustically irradiated medium is typically measured or estimated only globally by measuring the extinction length.

[0007] Losses caused by absorption and losses caused by scattering originate from different physical mechanisms. The associated characteristic lengths are more or less sensitive to the properties of the medium under study, thus affecting the detection of certain patient symptoms.

[0008] A paper titled "Multiple scattering of ultrasound in weakly inhomogeneous media: application to human soft tissue," authored by A. Aubry and A. Derode and published in J. Acoust. Soc. Am. 129, 225-233 (2011), discloses a method for the ultrasonic characterization of media, wherein the characteristic length l s and l a The components are measured independently. This method is based on distinguishing between single-scattering and multiple-scattering components. A single-scattering component consists of an echo that is scattered only once by a scatterer within the medium. This is used in ultrasound imaging to construct a reflectivity image of the medium, based on the relationship between echo time and the location of the scatterer in the medium under study. Multiple-scattering components correspond to waves that have undergone several scattering events before returning to the probe. Multiple scattering complicates the relationship between echo time and scatterer location. It significantly affects the construction of an image representative of the actual medium. The method described in this document relies on measuring a reflection matrix containing all impulse responses between each transducer of an ultrasound probe positioned towards the medium under study. By utilizing the differences in the statistical properties of single-scattering and multiple-scattering contributions, these components can be separated and the characteristic length l measured independently. s and l a However, this measurement is global and does not provide information about local variations in these two parameters.

[0009] Patent application publication WO 2021 / 023933 describes the determination of a focused reflection matrix, which, by considering the responses between virtual transducers, enables the local detection of the respective contributions of single and multiple scatterings, thereby allowing for a localized study of the propagation of ultrasound in a medium near the virtual transducers. However, such analysis does not provide information related to the absorption length and / or attenuation length and / or mean free path of scattering in the medium. Summary of the Invention

[0010] This disclosure aims to improve known ultrasonic detection methods, particularly to determine the local mean free path characteristics around one or more points within a medium.

[0011] According to a first aspect, this disclosure relates to an ultrasonic characterization method for a medium used in medical analysis, the method comprising the steps of: - generating a series of incident ultrasonic waves within a region of the medium via a transducer array, the series of incident ultrasonic waves forming an emitting base (i); and - measuring a canonical reflection matrix R defined between the emitting base (i) at an input and a receiving base (u) at an output. ui(t), where the coefficients of the canonical reflection matrix correspond to the signals generated by the ultrasonic waves received by the transducer and reflected within the medium; by using the canonical reflection matrix R ui (t) Focus on the spatial location r in = (x in The virtual input transducer (z) with spatial position r out = (x out Between the virtual output transducers of z), the focusing reflection matrix R of the medium is determined. xx (z), the two virtual transducers are located at the same depth z for the sound velocity model c0, and the focusing reflection matrix R xx The coefficients of (z) are written as follows: R xx (z) = [R( )]x in , x out Included around the point Within the area, virtual transducer r in and r out All lateral positions x in and x out A focusing basis (x) is formed at each depth z.

[0012] The method also includes: - focusing the reflection matrix Filtering is performed to obtain the single-scattering component representing the focused reflection matrix. Or, representing the multiple reflection matrix of the multiple scattering components of the focused reflection matrix. - By calculating the single reflection matrix R s The norm of (z) and the focused reflection matrix R xx The ratio of the norms of (z) determines the area around the point. The single scattering rate as a function of depth z within the region Alternatively, by calculating the reflection matrix R multiple times. M The norm of (z) and the focused reflection matrix R xx The ratio of the norms of (z) determines the area around the point. The multiple scattering rate as a function of depth z within the region .

[0013] Due to the effects of these arrangements, this method advantageously enables local probing of the medium to obtain local estimates of the single scattering rate and / or multiple scattering rate. This estimate represents the quality of the acquired measurements, and therefore the quality of the image that can be generated from these measurements.

[0014] These estimates are calculated from measurements acquired and recorded in the canonical reflection matrix. Therefore, they can be recalculated independently of the measurement acquisition stage, particularly by modifying various calculation parameters, enabling different ultrasound characterization analyses in real time or post-hoc.

[0015] These computations can be performed in the time domain or the frequency domain. Both computation modes provide essentially equivalent results, with reduced computational cost in the time domain and improved accuracy in the frequency domain.

[0016] According to various embodiments of the method, one and / or more of the following provisions may be additionally implemented.

[0017] According to a variant, filtering is performed by: - ​​determining a set of numerical focusing reflection matrices representing a single scattering in the medium. It is achieved by using a transducer array with a spatial position in a model medium having a sound velocity c0. The acoustic propagation between point scatterers (x, z) is calculated. in , x out , , and z are included around the point Within the specified region, the coefficients of the numerically focused reflection matrix are written as follows: - Determine a confocal base, which includes a reflection matrix that is focused by at least a portion of the numerical values. Several confocal matrices calculated by orthonormalization For depth z, k is from 1 to N k The index between, N k To represent depth A positive integer representing the number of matrices in the confocal basis; - will focus the reflection matrix. Projected onto a confocal matrix On the confocal basis, to obtain the single reflection matrix representing the single scattering component of the focused reflection matrix. Or, representing the multiple reflection matrix of the multiple scattering components of the focused reflection matrix. .

[0018] According to a variant: during the determination of the confocal group, in the form of Numerical reflection matrix expressed Arranged as a two-dimensional matrix defined for each depth z In the form of a two-dimensional matrix Perform singular value decomposition (SVD) according to the following expression:

[0019] in Two-dimensional matrix singular values, Two-dimensional matrix Singular vectors in the focusing basis (x), Two-dimensional matrix In having spatial location ( The singular vectors in the point scattering volume basis of z, with superscripts This represents the combined operation of conjugation and transpose.

[0020] According to a variation: for projection, the single reflection matrix Calculated as a confocal matrix in a confocal basis With the focused reflection matrix A linear combination of inner product weights at each depth The projection is obtained using the following formula:

[0021] in

[0022] Representation matrix and The inner product between, and the superscript This indicates a conjugate operation.

[0023] According to a variation: for projection, the multiple reflection matrix Calculated as a confocal matrix in a confocal basis With the focused reflection matrix A linear combination of inner product weights at each depth The projection is obtained using the following formula:

[0024] in

[0025] Representation matrix and The inner product between, and the superscript This indicates a conjugate operation.

[0026] According to a variant, the method further includes the following steps: - By identifying the single scattering rate Attenuation as a function of depth z, or by identifying the multiple scattering rate The increase as a function of depth z determines the area around the point. The mean free path of scattering within the region s .

[0027] Due to this regulation, at multiple points r within the medium pThe mean free path of scattering is obtained at l s The local estimate represents the single scattering behavior, and conversely, the multiple scattering behavior at various points in the medium, thus enabling the detection of abnormalities that may indicate the presence of pathological conditions in the identified region of the medium. Furthermore, the mean free path of this scattering can be periodically repeated. s This estimation is relevant for practitioners in order to monitor its evolution over time.

[0028] According to a variant, the method further includes: - targeting a set of points within the medium. Determine the mean free path of scattering The image.

[0029] According to a variant, the method further includes: - from the focused reflection matrix R xx The diagonal coefficients of (z) determine the confocal image I as a function of depth z. c (x,z), that is:

[0030] According to a variant, the method further includes: - Determining the area around the point The average confocal intensity is obtained by averaging the confocal image along its lateral dimension x within the region. :

[0031] Among the symbols This represents the average over the index variable.

[0032] According to a variant, the method further includes: - identifying the average confocal intensity The decay as a function of depth z determines the point. Extinction length l within the region ext .

[0033] According to a variant, the method further includes: - targeting a set of points within the medium. Determine the extinction length The image.

[0034] According to a variant, the method further includes: - determining the area around the point in the following manner Absorption length l within the region a : .

[0035] According to a variant, the method further includes: - targeting a set of points within the medium. Determine the absorption length The image.

[0036] According to a second aspect, this disclosure relates to an ultrasound characterization system for a medium used in medical analysis, configured to implement the method described above. The system according to the second aspect includes: - an array of transducers configured to generate a series of incident ultrasound waves within a region of the medium and to measure ultrasound backscatter waves from said region as a function of time; and - a processing unit associated with the transducer array, configured to implement the method according to the first aspect. Attached Figure Description

[0037] Further advantages and features of the above-described technology will become apparent from the following detailed description, given in a non-limiting manner and with reference to the accompanying drawings, in which: Figures 1(a) to 1(f) illustrate transmission and reception sequences for imaging and ultrasound characterization of a medium; Figure 2 illustrates conventional confocal imaging and matrix imaging; Figure 3 illustrates an example of an ultrasound characterization system for implementing the method according to the present disclosure; Figure 4 lists the definitions used in the method according to the present disclosure; Figure 5 is a flowchart of the ultrasound characterization method according to the present disclosure, in which single scattering rate or multiple scattering rate is determined; Figure 6 is a schematic diagram of an embodiment of the focusing reflection matrix filtering step of the method shown in Figure 5; Figure 7 illustrates the size of the region of interest and the resolution cell for calculation in various steps of the method; Figure 8 illustrates the application of the method of Figure 5 to a medium containing two Results of phantom imaging of cylindrical inclusions, each inclusion having different absorption and scattering properties; Figure 8(a) is a general schematic diagram showing the phantom positioned together with the transducer array; Figure 8(b) is an ultrasound image of a portion of the phantom with two measurement lines, the first line passing through the first inclusion and the second line passing through the second inclusion; Figure 8(c) shows the values ​​of the single scattering rate as a function of depth for the first and second lines of Figure 8(b); Figure 9 shows the results of applying the method of Figure 5 to an example of liver imaging; Figure 9(a) shows an ultrasound image of the liver with a transducer array positioned on the outer surface of the patient's body; Figure 9(b) is an ultrasound image of a portion of the medium in which the liver is located; and Figure 9(c) shows the values ​​of the single scattering rate as a function of depth in the portion of Figure 9(b).

[0038] In the various embodiments described with reference to the accompanying drawings, similar or identical elements have the same reference numerals unless otherwise stated. Detailed Implementation

[0039] In the following detailed description, only certain embodiments are described in detail to ensure clarity of presentation. These examples are not intended to limit the general scope arising from this disclosure.

[0040] The various embodiments and aspects described in this disclosure can be combined or simplified in a variety of ways. In particular, unless otherwise stated, the steps of the various methods can be repeated, interchanged, and / or performed in parallel.

[0041] This disclosure relates to methods and systems for ultrasound characterization of media, particularly suitable for medical imaging of living or non-living tissues. The media may be, for example, a non-homogeneous medium that is characterized to, for example, identify and / or characterize non-homogeneity. These characterization techniques are advantageously non-invasive to the media, and particularly preserve the properties and integrity of the media.

[0042] In conventional ultrasound imaging, the goal is typically to construct an image of the medium's reflectivity from the backscattered echoes from non-uniform sources within the medium. This is the principle behind ultrasound scanners used in medical imaging, which in particular enables the visualization of internal anatomical structures in individuals or animals. For simplicity, to construct an ultrasound image, the medium is assumed to be homogeneous with a constant sound propagation velocity c0.

[0043] Conventional ultrasound imaging methods typically use piezoelectric transducer arrays capable of independently or substantially independently transmitting and / or receiving ultrasound signals, each transducer located at position u along a support rod carrying the array. The medium-oriented transducer array allows for acoustic illumination of the medium in different ways and the construction of images representative of the medium. One conventional method involves acoustically illuminating the medium using focused emission via a technique called beamforming. This method involves applying an appropriate set of delays τ(u) based on a uniform velocity model c0 to the signal emitted by each transducer. in ,x in (x, z, c0) so that the lobes produced by each transducer have a spatial position (x, z, c0). in Interference occurs at the target focal point (z). Due to physical diffraction limitations, ultrasonic waves are emitted through the aperture of the ultrasonic probe and concentrated in a region commonly referred to as the "focal spot," with a lateral width δx.

[0044] To subsequently construct ultrasound images illustrating the properties of the studied medium, a digital focusing step is also performed during reception. The echo received by the array's transducers is brought back into phase through a time shift. Delay τ(u out , x out The delay (z, c0) is the same as that applied during launch, and the variable u out Specify the position of each transducer. During the launch phase, if the velocity model c0 used corresponds to the actual medium, then all signals at ballistic time t = z / c0 have positions (x, y, z). in Interference at the point (x, z). During reception, interference occurs from the same point (x, z). out = x inThe signal is coherently summed and interferometric at echo time t = 2z / c0. This summation produces the final result of the receiver focusing. This confocal method, involving dual focusing of transmission and reception, enables direct imaging of the reflectivity of the medium with lateral resolution δx and good contrast. However, this method is time-consuming because it requires physical focusing transmission at every point in the medium, or at least along every line of the constructed image representing the medium at a given depth.

[0045] A matrix-based ultrasound imaging method has been developed by the applicant in recent years. This method is based on constructing a canonical reflection matrix of the medium under study. This canonical reflection matrix is ​​typically determined experimentally. In other embodiments, the canonical reflection matrix can be determined by computation or numerical simulation reproducing the experiment.

[0046] The first method for creating this canonical reflection matrix involves sequentially emitting ultrasonic pulses from each transducer of the array, their positions determined by coordinates u. in The identification is shown in Figure 1(a). This produces a divergent cylindrical or spherical incident wave. The wave is reflected by a scatterer in the medium, and the backscattered field is measured as a function of time for each transducer, as shown in Figure 1(b). By repeating this operation for each transducer used sequentially as a source, the canonical reflection matrix R is determined in the transducer basis. uu (t) = [R(u) out , u in [,t)], containing all impulse responses R(u) between each pair of transducers. out , u in Therefore, this matrix contains a wealth of information describing the medium under study. However, this method assumes that the medium remains stationary throughout the measurement duration, which is challenging, especially for medical examinations performed on living patients. Furthermore, the recorded signal exhibits a poor signal-to-noise ratio because the medium is acoustically irradiated by a single transducer.

[0047] A second method for constructing this canonical reflection matrix involves acoustically irradiating the medium with a series of plane waves. This method overcomes most of the limitations inherent in the aforementioned methods. Figure 1(c) illustrates the principle of plane wave irradiation. A delay law τ′ is applied to each transmitted signal to generate a tilt angle θ relative to the transducer array. in The wavefront. During reception, as shown in Figure 1(d), for each incident plane wave θ in The field R(u) backscattered by the medium out , θ in , t) is located at position u out All sensor measurements. The set of these responses forms the canonical reflection matrix R. uθ (t) = [R(u) out , θ in[, t). The aforementioned dual-focusing method can then be numerically implemented by time-shifting the measurement signal before coherent summation. This method leads to ultrafast imaging and elastography, and is described in the following document: “Coherent plane-wave compounding for very high frame rate ultrasonography and transientelastography”, G. Montaldo et al., IEEE Transactions on Ultrasound, Ferroelect. Freq. Control 56 489-506, 2009.

[0048] A third alternative for creating this canonical reflection matrix involves acoustically irradiating the medium with a divergent wave basis, as shown in Figures 1(e) and 1(f), which allows for a wider illumination of the sound field than using plane waves. This technique is particularly useful for super-resolution imaging and is described in the following document: “Ultrafast imaging of the heart using Circular Wave Synthetic Imaging with Phased Arrays”, Couade et al., IEEE International Ultrasonics Symposium (2009).

[0049] Therefore, conventional imaging includes emission and reception at each image pixel (x in = x out Double focus is performed at the same focal point as shown in Figure 2(a).

[0050] In contrast, matrix imaging involves decoupling the transmission and reception focus for a given echo time t, as shown in Figure 2(b). The focused reflection matrix R is thus determined. xx ( ) = [R(x in ,x out , z, The focused reflection matrix includes the response between the virtual input transducer and the virtual output transducer, which each have a spatial position r. in = (x) in ,z) and r out = (x) outThe depth z corresponds to the composite focal spot at different spatial locations of the transmitter and receiver. The expected depth z is determined by the echo time t: z =t / 2. Here, Indicates the time in the focusing base, such that .time The origin corresponds to the moment when the virtual source emits the ultrasonic pulse. The focused reflection matrix is ​​obtained, for example, from the canonical reflection matrix R. ui (t) Beamforming focusing is used to obtain the matrix. This matrix provides access to more feature information about the medium under study than that obtained using conventional confocal imaging modes. The focused reflection matrix can be synthesized in the time domain using a delay law, or in the frequency domain by applying an appropriate phase shift.

[0051] The focused reflection matrix is ​​specifically described in the following documents: “Reflection Matrix Approach for Quantitative Imaging of Scattering Media”, William Lambert et al., Phys. Rev. X10, 021048, (2020), and subsequently in “Ultrasound Matrix Imaging - Part I: The Focused Reflection Matrix, the F-Factor and the Role of Multiple Scattering”, William Lambert et al., IEEE Trans. Med. Imag. 41, 3907-3920, (2022).

[0052] In these publications, during ballistic time (t=2z / c0) Consider the focusing reflection matrix R between the virtual transducers xx (z, Although matrix R xx (z, The diagonal coefficient (x) of (=0) out = x in This allows for the construction of a synthetic confocal image at depth z, but its off-diagonal coefficients provide information about aberration effects and multiple scattering that may degrade the image quality.

[0053] In patent applications WO2020 / 016250 and WO2021 / 023933, the focused reflection matrix was studied in order to quantify the multiple scattering rate in particular.

[0054] In those public and applied-for documents, by examining the confocal intensity |R(x) in =x out ,z, )| 2 With off-axis strength |R(x) in x out ,z, )| 2 The ratio between single and multiple scattering rates is used to determine the multiple scattering rate; the former accumulates both single and multiple scattering contributions, while the latter includes only the multiple scattering contribution. However, this method is only qualitative because the focal spot exhibits sidelobes due to diffraction, and the non-negligible portion of the single scattering contribution also appears in the off-diagonal coefficients of the focused reflection matrix. Furthermore, these methods do not utilize single or multiple scattering rates to quantitatively measure the attenuation length, absorption length, and mean free path of the scattering medium. To achieve quantitative measurements of these parameters, diffraction effects must be more precisely accounted for in order to probe the spatial evolution of single and / or multiple scattering rates.

[0055] Figure 3 illustrates an example of an ultrasound imaging system 1 according to the present disclosure for ultrasound imaging of a medium, such as a non-homogeneous medium M. This system and method enable the formation of ultrasound images of at least a portion of the medium, corresponding to the region of interest or field of view.

[0056] System 1 includes: - a detection device 20 or probe 20, - a processing unit 30 for calculating an image based on signals received from the probe 20, - a control panel 40 connected to the processing unit 30, the control panel including, for example, buttons 41 and a touchpad 42, - a display device 50 for visualizing the image and various components or measurements.

[0057] The probe 20 is connected to the processing unit 30 via cable 21 or wireless connection, and is capable of emitting and receiving ultrasonic waves W into and from the medium M, the ultrasonic waves being generated by the reflection of the emitted ultrasonic waves by scattering particles or scatterers within the medium.

[0058] The probe 20 may include an array 10 comprising a plurality of transducers 11. The array 10 may be, for example, a linear, curved, two-dimensional, or matrix array. The transducers 11 are capable of converting electrical signals into vibrations and vice versa. The transducers 11 may be, for example, piezoelectric ultrasonic transducers, and may take the form of rigid rods, placed in direct or indirect contact with the outer surface of the medium M to acoustically couple with the medium and vibrate to emit and receive ultrasonic waves W. The array 10 of transducers 11 of the probe 20 is associated with a processing unit 30. The array 10 may include one hundred or more transducers 11.

[0059] The processing unit 30 may include a housing 31, which includes receiving means for amplifying and / or filtering signals received from the probe 20, and converters, including analog-to-digital and digital-to-analog converters, for converting the signals into data representing the signals. The data may be stored in the memory of the processing unit 30 and / or processed directly to calculate intermediate data (beamforming or other data). The processing unit 30 may implement any known method for constructing an image based on the signal data received from the probe 20, such as beamforming.

[0060] The images computed can be: - images of the medium (B-mode images), typically in grayscale, used to visualize organs within the medium, and / or - images showing velocity or flow within the medium (color images), such as for visualizing blood vessels within the medium, and / or - images showing the mechanical properties (elasticity) of the medium, such as for identifying tumors within the medium.

[0061] The "connection" or "link" between the detection device 20, the processing unit 30, and the display device 50 encompasses any type of wired connection, electrical or optical connection, or connection using any protocol such as WiFi. TM Or Bluetooth TM Wireless connections of any type or other protocol. These connections may be unidirectional or bidirectional. The associated display device 50 may be of any type, such as a touchscreen or non-touchscreen, connected or independent.

[0062] Display device 50 is a screen that enables visualization of images calculated by processing unit 30. Display device 50 may also display other information, such as image scale, configuration information for calculations or processing, or any measurement or auxiliary information. Screen 50 may be mounted on support arm 51 to provide improved positioning for the user. Screen 50 is typically large, for example, at least 20 inches, to provide improved visibility for the user.

[0063] The control panel 40 may form part of the system housing, the part including a panel housing having a substantially planar surface 40a inclined toward the user to allow for one-handed operation. As shown in FIG3, the control panel 40 may include a control screen 49 for displaying various configuration information.

[0064] The processing unit 30 is configured to perform calculation and / or processing steps, particularly for performing method steps according to this disclosure. Conventionally, as shown in FIG4, an array 10 of transducers 11 on the surface of a medium M is illustrated, defining a spatial coordinate system for the medium M with a first axis X and a second axis Z perpendicular to the first axis. For simplicity, the first axis X corresponds to the transverse direction in which the transducers 11 are linearly aligned in this example, and the second axis Z corresponds to the depth of the array 10 of transducers 11 within the medium M relative to the depth of the array 10 of transducers 11. This definition can be adapted to the context and can be extended, for example, to a three-axis spatial coordinate system in the case of a two-dimensional array 10, a polar coordinate system in the case of a curved array 10, or any other coordinate system adapted to and / or dependent on the structure and shape of the ultrasonic transducer array 10. Therefore, in the remainder of this disclosure, the Cartesian coordinate system XZ corresponding to the linear probe 20 will be used for simplification, although those skilled in the art will readily generalize and apply the results to any type of coordinate system.

[0065] In the remainder of this disclosure, with reference to the array 10 of transducers 11 for transmitting and receiving, it should be understood that, more generally, several transducer arrays may be used simultaneously. Transducers 11 may operate as transmitters and subsequently as receivers, or some transducers may operate only as transmitters while others operate only as receivers. Similarly, array 10 may comprise from 1 to N transducers 11, of the same or different types.

[0066] The array 10 of transducers 11 can operate, for example, as both a transmitter and a receiver, or it can consist of several transducer subarrays, some dedicated to transmitting ultrasonic waves and others dedicated to receiving ultrasonic waves. The term transducer array means at least one transducer, an aligned or unaligned sequence of transducers, a two-dimensional distribution of transducers (e.g., a transducer matrix), or any spatial distribution of transducers.

[0067] In the context of this disclosure, where references are made to computational or processing steps for implementing particular method steps, each computational or processing step may be implemented by software, hardware, firmware, microcode, or any suitable combination of these or related technologies. When using software, each computational or processing step may be implemented by computer program instructions or code, which may, for example, be interpreted or executed. Such instructions may be stored in or transmitted to a computer-readable storage medium (or processing unit) and / or executed by a computer (or processing unit) to implement these computational or processing steps.

[0068] Focused Reflection Matrix Analysis at a Point in a Medium This disclosure describes methods and systems for ultrasonic characterization of media. In practice, it is assumed that the medium is non-homogeneous. These methods and systems are based on the definitions shown in Figure 4: Within the medium, the following is defined: - Spatial position r inThe first point P1, and - spatial location r out The second point, P2.

[0069] Spatial position r in and r out These elements are shown in bold to indicate that they are position vectors taken in the medium space coordinate system (X, Z). Other representations and definitions of point positions are possible and readily accessible to those skilled in the art of ultrasound.

[0070] In this disclosure, the first point P1 has a value denoted as x. in The horizontal position of the second point P2. Let x be the value of the second point. out The lateral position. Two points P1 and P2 have the same expected depth z. in = z out Also denoted as z, it is controlled by the considered echo time t. Therefore, the spatial positions of points P1 and P2 are r and r, respectively. in = (x) in ,z) and r out = (x) out ,z).

[0071] The two points P1 and P2 are chosen to be relatively close to each other, i.e. a few millimeters apart, such as twenty (20) millimeters or less.

[0072] As shown in Figure 5, the ultrasonic characterization method S100 implemented by the processing unit 30 of system 40 includes the following steps: - generating a series of incident ultrasonic waves US in the region of the medium through the array 10 of transducers 11. in The series of incident ultrasonic waves form a transmitting base i; and - the canonical reflection matrix R defined by S120 between the transmitting base i at the input and the receiving base u at the output is measured or constructed. ui (t), where the coefficients of the canonical reflection matrix correspond to the signal generated by the ultrasonic waves received by the transducer and reflected by the scatterer within the medium; - through the canonical reflection matrix R ui (t) Focusing, determining S130 including spatial location r in Virtual input transducer TV in With spatial position r out Virtual output transducer TV out The focusing reflection matrix R of the medium response between xx (z).

[0073] The obtained canonical reflection matrix R ui(t) can be a "real" matrix, that is, composed of real coefficients in the time domain, with each transducer recording a real electrical signal. Alternatively, the matrix can be a "complex" matrix, that is, including complex values, for example in demodulation cases used for in-phase and quadrature beamforming ("IQ beamforming").

[0074] Focused reflection matrix R xx It can be expressed in different ways. In the expression disclosed herein, the spatial position is taken as (x in The first point P1 of (x, z) can be used as a reference. These expressions can also be relative to the spatial position (x, z). out The second point P2 is established relative to the midpoint between P1 and P2, and its spatial position is ((x, z). in + x out () / 2, z), or relative to any specified reference point. Those skilled in the art will be able to make the necessary variable substitutions in the presented expression.

[0075] By using the canonical reflection matrix R ui (t) Focusing to calculate the medium response.

[0076] Focused reflection matrix R xx The response of (z) corresponds to a transverse position x in the medium at the expected depth z and echo time t, calculated for the assumed sound velocity c0. in and x out The sound pressure field between all points. In other words, the focusing reflection matrix R xx (z) is defined as: R xx (z) = [R(x) in , x out The depth z and sound velocity c0 in the parameter medium affect the delay law used in the focusing process.

[0077] The emission basis i at the input is, for example, a basis of a wave generated by one of the transducers 11 of the array 10, or a basis of a plane wave with an angle tilt θ relative to the axis X, or a basis of a virtual source, as described in the preceding figures 2(a) to 2(f).

[0078] The receiving base u is, for example, the base of transducer 11. Alternatively, another receiving base, such as a frequency or spectrum base, can be used during reception.

[0079] Therefore, the ultrasound generation step S110 is understood to occur between the transmitting base i and the receiving base u. Thus, this ultrasound generation step is defined for all types of ultrasound, whether focused or unfocused, such as plane waves.

[0080] In measurement step S120, the normalized reflection matrix R is... ui(t) is defined between the transmitting basis i at the input and the receiving basis u at the output. This matrix contains all the time responses of the medium, given by spatial coordinates u at time t. out Each transducer 11 measures and is used for each transmission i in Elements specified by the index "in" indicate transmission (i.e., input), and elements specified by the index "out" indicate reception (i.e., output). This canonical matrix can also be recorded and / or stored, for example in the memory of the processing unit, or on any other medium that enables permanent or temporary storage, whether movable or immovable.

[0081] In step S130, the focused reflection matrix R xx (z) can be determined or calculated in the following way: - Based on the canonical reflection matrix R ui The input focusing process of (t) uses a wave at the transmitting base i and a virtual input transducer TV. in Forward propagation time between, and in spatial location r in An input focal spot is generated around the first point P1, the input focal spot corresponding to the virtual input transducer TV. in - Based on the canonical reflection matrix R ui The output focusing process of (t) uses a wave in the virtual output transducer TV. out The return propagation time between the transducer and the receiving base u, and at spatial location r out The second point P2 generates an output focal spot, which corresponds to the virtual output transducer TV. out .

[0082] These input and output focusing processes form an input-output focusing process, referred to as a focusing process or more simply as focusing in the remainder of this disclosure.

[0083] In other words, in this ultrasonic characterization method, the virtual input transducer TV in Corresponding to the spatial position r in the medium in The ultrasound "virtual source", virtual output transducer TV out Corresponding to the spatial position r in the medium out An ultrasonic "virtual sensor". The virtual source and virtual sensor are connected by the difference in their spatial positions Δr=r. out -r in They are separated in space. In this case, they are separated only along the transverse axis, by Δx = x out -x inTheir expected depth is the parameter z used in the focusing law for the sound velocity model c0. Their actual depth is determined by the axial position (depth) of the isochronous volume, i.e., by the echo time t and the distribution of the sound velocity c(r) in the medium. The lateral dimension of the virtual transducer is determined by the focal spot produced by focusing at this actual depth.

[0084] In addition, the focused reflection matrix R xx (z) can be determined or computed: - or in the time domain, in which case it can be explicitly expressed using time parameters. This is represented as R. xx (z, ), and in fact, the focused reflection matrix R xx The data for (z, t) are calculated between two predetermined times; or in the frequency domain, in which case it can be explicitly expressed using the angular frequency parameter. This indicates that the angular frequency parameter pass Corresponding to frequency f, denoted as R xx (z, ), and in fact, the focused reflection matrix R xx (z, The data is located at both angular frequencies within the frequency bandwidth. Calculations between, for example, at lower angular frequencies With higher angular frequency Between, and the center angular frequency .

[0085] Therefore, subsequent calculations using this method can be performed in the time domain or the frequency domain.

[0086] In the first case of time-domain computation, the virtual input transducer TV in With virtual output transducer TV out The focusing reflection matrix R of the medium between xx (z, The focused reflection matrix R is obtained by calculating the focusing through beamforming of the transmitted and received beams. xx (z, The coefficient of ) can be determined in the following way: (1) Where: N in N represents the number of elements in the emitter base (i). out R is the number of elements in the receiving basis (u). ui (t) is the confocal reflection matrix, where R(u) out i in , t) is the confocal reflection matrix R ui The elements of (t) are those with spatial position u. out The transducer is indexed as i in the transmitter base (i).in After the launch, and recorded at time t, a delay time τ was applied to it. in and τ out ; and It is a predefined apodization factor, for example, to maintain a constant numerical aperture for both transmission and reception; For each incident wave Arrival spatial location is The first focus is the expected propagation time in the model medium with the speed of sound c0. From the perspective of spatial location The second focus to the position is The expected propagation time of the reflected wave from the transducer.

[0087] These delay times τ in and τ out Propagation time is typically calculated by those skilled in the art based on established sound velocity models. Relatively simplified assumptions include the assumption of a homogeneous medium with a constant sound velocity c0. In this case, the propagation time is obtained directly from the distance between the probe transducer and the virtual transducer. Therefore, these delay calculations depend on the wave type, the assumed sound velocity, and the geometry of the transducer array.

[0088] The number of elements N of the emitter group in For example, greater than or equal to 1, advantageously greater than or equal to 2. The number of elements N of the acceptor base. out For example, greater than or equal to 2.

[0089] Therefore, the above beamforming formula is recorded in the canonical reflection matrix R. ui The double summation of the time response corresponds to the first summation of the transmit focusing with respect to the transmit basis i, and the second summation of the receive focusing with respect to the receive basis u. This calculation is performed with respect to the spatial coordinates of two points P1 and P2, i.e., their respective expected spatial positions r. in = (x) in ,z) and r out = (x out Therefore, the result of this beamforming formula is for these two spatial coordinates (r, z). in ,r out ) or these two horizontal positions x in ,x out The time signal.

[0090] Finally, the focused reflection matrix R expressed in the time domain xx (z, It can be converted into the focused reflection matrix R in the frequency domain through Fourier transform. xx (z, ), that is, through: (2) The Fourier transform can be implemented by any type of discrete Fourier transform, whether or not it is normalized.

[0091] In the second case of frequency domain computation, the canonical reflection matrix R, expressed in the time domain, is composed of the signal received by the transducer. ui (t) can be transformed into the canonical reflection matrix R in the frequency domain by Fourier transform. ui ( ), that is, through: (3) The Fourier transform can be implemented by any type of discrete Fourier transform, whether or not it is normalized.

[0092] Therefore, the focusing reflection matrix R of the medium xx (z) or R xx (z, The focusing can be obtained through the following matrix calculation, which is essentially equivalent to the time-domain beamforming described above, i.e., through the following matrix product: (4) Where the matrix To normalize the reflection matrix Fourier transform, matrix From the receiving base (u) to the depth z and angular frequency The transfer matrix of the focusing basis (x) at point, matrix From the emitter (i) to the depth z and angular frequency The transfer matrix of the focusing basis (x) at point, sign These represent matrix operations for conjugate and conjugate transpose, respectively.

[0093] The local determination of single scattering rate / multiple scattering rate according to the method S100 of this disclosure further includes the following steps: - for the focused reflection matrix conduct Filtering S140 obtains a single-scattering matrix representing the single-scattering component of the focused reflection matrix. Or, representing the multiple reflection matrix of the multiple scattering components of the focused reflection matrix. - By calculating the single reflection matrix R s The norm of (z) and the focused reflection matrix R xx The ratio of the norms of (z) determines the S150 around the point. Within the region, it is a function of depth z Single scattering rate Alternatively, by calculating the reflection matrix R multiple times. M The norm of (z) and the focused reflection matrix R xx The ratio of the norms of (z) determines the area around the point. Within the region, it is a function of depth z Multiple scattering rate .

[0094] The single scattering matrix and the multiple scattering matrix are complementary to each other, as shown in relations (14) and (14'). The single scattering rate and the multiple scattering rate are also complementary to each other, as explained in relations (5') and (8'). Therefore, the calculation can be performed according to one of two alternatives to the filtering step S140 and the determination step S150. These different possible calculations are detailed below.

[0095] Due to the effects of these arrangements, the method advantageously enables local probing of the medium to obtain local estimates of the single scattering rate and / or multiple scattering rate.

[0096] These quantities represent the level of multiple scattering in the final constructed ultrasound image and therefore its reliability. Generally, a multiple scattering rate greater than 0.5 indicates that the single scattering assumption upon which conventional ultrasound imaging is based is invalid, and therefore the resulting image is not a correct estimate of the medium's reflectivity. Conversely, a multiple scattering rate less than 0.5 (a single scattering rate greater than 0.5) indicates that the single scattering assumption of conventional ultrasound imaging is valid.

[0097] Furthermore, these estimates are calculated from measurements performed and recorded in the canonical reflection matrix. These estimates can be recalculated at any time by modifying certain calculation parameters, enabling updated ultrasound characterization analysis in real time or ex-post, such as offline.

[0098] In particular, in step S150, the single scattering rate It can be calculated using the following ratio: (5) Among them: For matrix The Frobenius norm, Let X be the trace of matrix X. Let M be the conjugate transpose of matrix M.

[0099] Therefore, single scattering rate From a single reflection matrix and focused reflection matrix get.

[0100] This single scattering rate The single reflection matrix can be calculated, either directly or from the time domain. and focused reflection matrix R xx (z), and the aforementioned formula can be expressed as: (6) - or the single reflection matrix calculated from the frequency domain and focused reflection matrix R xx (z), and the aforementioned formula can be expressed as: (7) Then, the single scattering rate is determined over the frequency bandwidth of interest, for example at lower angular frequencies. With higher angular frequency between.

[0101] Conversely, in step S150, the single scattering rate It can be calculated using the following ratio: (5') Wherein: For matrix The Frobenius norm, Let X be the trace of matrix X. Let M be the conjugate transpose of matrix M.

[0102] Therefore, single scattering rate It can also be obtained from multiple reflection matrices and focused reflection matrix get.

[0103] This calculation can also be performed in the time domain or frequency domain, as previously done for calculating the single scattering rate based on multiple reflection matrices. Same.

[0104] Therefore, single scattering rate From a single reflection matrix and focused reflection matrix get.

[0105] In step S150, multiple scattering rate Conversely, the calculation is as follows: (8) Among them: For matrix The Frobenius norm, Let X be the trace of matrix X. Let M be the conjugate transpose of matrix M.

[0106] Therefore, multiple scattering rate From multiple reflection matrices and focused reflection matrix get.

[0107] This calculation can be performed in the time domain or the frequency domain, as previously described.

[0108] Conversely, in step S150, the multiple scattering rate It can be calculated using the following ratio: (8') Wherein: For matrix The Frobenius norm, Let X be the trace of matrix X. Let M be the conjugate transpose of matrix M.

[0109] Therefore, multiple scattering rate From a single reflection matrix and focused reflection matrix get.

[0110] These calculations can be performed in the time domain or the frequency domain, as previously described.

[0111] Therefore, single scattering rate It can be determined by calculating the single scattering matrix or the multiple scattering matrix. Similarly, the multiple scattering rate... It can be determined by calculating the single scattering matrix or the multiple scattering matrix.

[0112] Therefore, according to the method of this disclosure, the depth z-space position in medium M is The single scattering rate or multiple scattering rate is locally determined at a given point.

[0113] However, the focused reflection matrix must be filtered before calculating the single scattering rate or multiple scattering rate. The following paragraphs of this disclosure explain how such filtering is performed.

[0114] According to an embodiment of the method of this disclosure, the filtering of the focused reflection matrix, filtering S140, can be performed according to the following steps, as shown in FIG6: - Determine a set of numerical focusing reflection matrices for S141. It represents a single scattering in the medium, through a transducer array with a spatial position in a model medium having a sound velocity c0. The acoustic propagation between point scatterers (x, z) is calculated. in , x out , , and z are included around the point Within the specified region, the coefficients of the numerically focused reflection matrix can be written as follows: .

[0115] - S142 is determined to include several confocal matrices. confocal group It focuses the reflection matrix by at least a portion of the numerical values. Orthogonal normalization is used for calculation, for depth z and angular frequency k is from 1 to N k The index between, N k To represent depth The number of positive integers in the confocal basis, - Focus the reflection matrix Projector S143 To confocal matrix On the confocal basis, to obtain the single reflection matrix representing the single scattering component of the focused reflection matrix. Or, representing the multiple reflection matrix of the multiple scattering components of the focused reflection matrix. .

[0116] Steps S141, S142, and S143 are also performed on the time-domain data or on the bandwidth in frequency-domain mode, for example, at lower angular frequencies. With higher angular frequency Between, and the center angular frequency .

[0117] In step S141, this set of numerical focusing reflection matrices The ultrasonic mathematical model corresponding to the medium.

[0118] In step S142, several confocal matrices of the model are constructed. Confocal groups, limited to N k A confocal matrix, which enables the focused reflection matrix measured in the real medium by the array 10 of transducers 11 through the projection pair performed in step S143. Perform truncation or filtering.

[0119] The filtering arrangement of these focused reflection matrices enables the acquisition of local characteristics within the medium, particularly in medium characterization methods used to determine single or multiple scattering rates, within a confocal base. Furthermore, the values ​​determined by such filtering are advantageously more accurate than those in prior methods. This filtering method is also easier to implement.

[0120] According to a first embodiment of the method, in step S142, the determination of the confocal basis can be performed by Gram-Schmidt orthogonal normalization calculation.

[0121] According to a second embodiment of the method, determining the confocal base in step S142 can be performed as follows: in the form of Numerical reflection matrix expressed Arranged as a two-dimensional matrix defined for each depth z Form; for two-dimensional matrices Perform singular value decomposition M according to the following expression: (9) Among them: Two-dimensional matrix singular values, Two-dimensional matrix Singular vectors in the focusing basis (x), Two-dimensional matrix In having spatial location ( The singular vectors in the point scattering volume basis of z), and the superscript This represents the combined operation of transpose and conjugation.

[0122] Because of the effectiveness of these methods in determining the arrangement of the confocal basis through singular value decomposition by arranging the numerical matrix in two dimensions, more accurate filtering of the focused reflection matrix is ​​achieved. Furthermore, rank N... k This allows for the optimal determination of the filtering dimension. Such computations are also faster and easier to implement than the Gram-Schmidt method.

[0123] Figure 7 illustrates examples of dimensions used in the calculations at various steps of this method, which uses an array 10 of transducers 11 with dimension D in the lateral direction x for receiving a basis u, and seeks to characterize a spatial location r. p The area surrounding the point. The lateral analysis region has lateral dimensions. Each resolvable cell along the line at depth z has a lateral dimension. and vertical dimension Distinguish the lateral dimension of the unit. It depends on the depth z.

[0124] During step S142, when the confocal basis is determined, each two-dimensional matrix rank Then, the number of resolvable cells contained within the region is fixed, i.e., calculated as follows: (10) of which For the surrounding point The horizontal dimension of the region, Let z be the lateral dimension of the resolvable cell at depth z.

[0125] These quantities, when calculated in the frequency domain, depend on the angular frequency. .

[0126] The lateral dimension of the resolvable cell is estimated, for example, through the following types of physical models: (11) of which For the speed of sound Lower angular frequency The wavelength at that point, and D is the lateral dimension of the array 10 of transducers 11.

[0127] Therefore, the singular value decomposition M-procedure enables the determination of the numerical reflection matrix. The singular vectors in the focusing basis (x) are the dominant elements in constructing a filter matrix that contains only single or multiple scattering components.

[0128] In step S143, the projection step enables the construction to be performed, and the projection is performed according to a first alternative scheme for calculating the single scattering matrix in the following manner: single reflection matrix The calculation is performed as a confocal basis confocal matrix. With the focused reflection matrix A linear combination of inner product weights at each depth The projection is obtained using the following formula: (12) of which (13) represents a matrix and The inner product between, and the superscript This indicates a conjugate operation.

[0129] In step S143, the projection step enables the construction to be performed, and the projection is performed according to a second alternative scheme for calculating the multiple scattering matrix in the following manner: multiple reflection matrix The calculation is performed as a confocal basis confocal matrix. With the focused reflection matrix A linear combination of inner product weights at each depth The projection is obtained using the following formula: (12') of which (13) represents a matrix and The inner product between, and the superscript This indicates a conjugate operation.

[0130] You can convert a single scattering matrix to a multiple scattering matrix, or vice versa.

[0131] Therefore, the multiple scattering matrix It can be obtained from the focused reflection matrix Subtract the single scattering matrix To obtain; that is, through: (14) Therefore, the single scattering matrix It can be obtained from the focused reflection matrix Subtract the multiple scattering matrix To obtain; that is, through: (14') Therefore, it is possible to: - calculate the single reflection matrix directly by projection of formula (12), or - calculate the multiple reflection matrix directly by projection of formula (12'), or - calculate the single reflection matrix by projection of formula (12) and obtain the multiple reflection matrix by subtraction of formula (14), or - calculate the multiple reflection matrix by projection of formula (12') and obtain the single reflection matrix by subtraction of formula (14').

[0132] These different variations then make it possible to determine through various equivalent computational paths. Single scattering rate and / or multiple scattering rate .

[0133] Scattering mean free path l sThe applicant has identified the point on medium M. Determined local single scattering rate With local scattering mean free path The relationship between values.

[0134] In particular, single scattering rate It is a decreasing function of depth z.

[0135] For example, the decreasing function can be a linear or exponential function that decreases as a function of depth z within the scope of interest. Other decreasing function models can also be applied.

[0136] For linear function types, for example, we can have: (15) For exponential type functions, we can have: (16) of which The predetermined attenuation parameter is used. This attenuation parameter is typically between half (0.5) and 1. Furthermore, it depends on the characteristics of the probe 20, such as its geometry or the directionality of the transducer 11 forming the probe 20.

[0137] In particular, the inventors' analysis revealed that, surprisingly, the attenuation parameter is essentially optimal for two-thirds of the values: .

[0138] Furthermore, the complementarity relationship can be applied to the multiple scattering rate. Conversely, the mean free path of the scattering is determined. : (17) In the aforementioned relationship. In this case, the multiple scattering rate It is an increasing function that is a function of depth z.

[0139] Therefore, the ultrasonic characterization method for medium M described in this disclosure may optionally include the following additional steps, as shown in Figure 5: - By identifying the single scattering rate As depth decreases, or by identifying the multiple scattering rate As depth increases, the S160 surrounding point is determined. The mean free path of scattering within the region s .

[0140] This step includes identifying or determining the single scattering rate obtained by the decreasing function curve as a function of depth z. The best fit of the value is used to determine the single scattering rate. The same fitting procedure applies to multiple scattering rates. .

[0141] Examples of decreasing function curves are shown as dashed lines in Figure 8(c) and Figure 9(c) below, which is described in more detail. In these cases, the decreasing function curve is quasi-linear and best approximates the location of the single scattering rate determination point as a function of depth z.

[0142] Scattering mean free path l s The calculation is around the point The local estimation. Therefore, the ultrasonic characterization method of medium M determines the corresponding set of points within the medium. Scattering mean free path Images or diagrams.

[0143] This plot or image of the mean free path of scattering in a medium provides a quantitative characterization of the physical properties of the medium within its volume. The mean free path of scattering is a parameter indicating pathology within the medium. This mapping enables the localization of said pathology.

[0144] According to an embodiment of the ultrasound characterization method of this disclosure, the confocal image and average confocal intensity further include the following steps: - Determine confocal image I c (x,z) as a function of depth z, from the focused reflection matrix R xx The diagonal coefficients of (z) are determined by the following calculation: (18) Such confocal images correspond to conventional ultrasound images that are free from any multiple scattering effects.

[0145] Using a confocal image, the following steps can then be performed: - Determine the surrounding point The average confocal intensity is obtained by averaging the confocal image along its lateral dimension x within the region. That is, through the following calculation: (19) Among them, the symbols This represents the average over the index variable, i.e., the horizontal dimension x.

[0146] Extinction length l ext and absorption length l a According to an embodiment of the ultrasonic characterization method of this disclosure, as shown in Figure 5, once the average confocal intensity is calculated, the optional step can be performed: - By identifying the average confocal intensity The decay as a function of depth z determines the S170 around the point. within the area Extinction length l ext .

[0147] According to the first variant, Extinction length l ext By identifying the confocal intensity as a function of depth z It is determined by the average linear decrease over the region of interest.

[0148] According to the second variant, the extinction length l ext By identifying confocal intensity as a function of depth The exponential decrease in the average over the region is determined by the following formula: (20) where I0 is a constant independent of depth z.

[0149] The identification method described for determining the mean free path of scattering is applied in a similar manner. A set of values ​​for the confocal intensity at different depths z allows for the inference of the area around the point. extinction length l ext .

[0150] Similarly, a set of points within medium M can be generated. extinction length The figure corresponds to an image of the medium characterization parameters.

[0151] As shown in Figure 5, the optional step can then be performed: - Determine the S180 surrounding point in the following manner. within the area Absorption length l a : (21) It can also generate a set of points within medium M. absorption length The figure corresponds to an image of the medium characterization parameters.

[0152] Therefore, through the various calculations described above, the absorption loss and scattering loss of medium M can be distinguished during acoustic illumination. This allows for a more accurate characterization of medium M in a local manner than prior art methods that only provide global estimates of these parameters. In particular, the formulas presented in this disclosure are valid over a wide depth range, meaning they are valid in the near field, not just the far field.

[0153] Figure 8 shows the results of applying the method of this disclosure to a phantom containing two cylindrical inclusions with different reflectivity values ​​and different single-scattering characteristics. Figure 8(a) shows a general schematic diagram of the phantom obtained using a transducer (11) of an array (10) positioned on the outer surface of the phantom on the left side of the figure to emit ultrasonic waves to the right (i.e., in the depth direction z). The region of interest surrounds the two inclusions I1, I2.

[0154] System (1) generates an ultrasound image of the region of interest. This ultrasound image is shown in Figure 8(b). The image shows that the first inclusion I1 is on average brighter than the second inclusion I2. Two measurement lines are then defined along the depth direction: a first line L1 passing through the first inclusion and a second line L2 passing through the second inclusion.

[0155] The method according to this disclosure is applied to a set of points (with spatial location r) along each of these lines. p (Points). For each of these points, a region, such as a lateral point region, is used to apply the method. Figure 8(c) shows the value V1 of the single scattering rate obtained as a function of depth along the first line L1 and the value V2 of the single scattering rate obtained as a function of depth along the second line L2. Therefore, this method enables the accurate identification of local values ​​of the single scattering rate.

[0156] According to the method of this disclosure, it is then possible to identify a decreasing function F1 for the depth range of a first inclusion and a decreasing function F2 for the depth range of a second inclusion. These functions exhibit different attenuation behaviors, representing the so-called scattering mean free path l of each inclusion. s By using theoretical curves By fitting experimental points, this method thus enables sufficient differentiation of bright inclusions (V1, l). s ~32 mm) and dark inclusions (V2, l) s The local mean free path was measured in ~80 mm, and both inclusions were present in the medium.

[0157] Results on the liver Figure 9 shows the results of applying the method according to this disclosure to a mammalian liver irradiated by the system (1). Figure 9(a) shows an ultrasound image of a liver having a transducer (11) with an array (10) positioned on the outer surface of the subject's body (left side of the figure) to emit ultrasound waves to the right (i.e., in the depth direction z). The region of interest surrounds the subject's liver.

[0158] An ultrasound image of the region of interest is generated by system (1). This ultrasound image is shown in Figure 9(b), which enables visualization of the liver.

[0159] The method according to this disclosure is applied to a set of points (with spatial position r) along a line extending in the depth direction z. p (Points). For each of these points, a region, such as a lateral point region (in the x-direction), is used to apply the method. Figure 9(c) shows the value V of the single scattering rate obtained as a function of depth along the line under consideration. Thus, this method enables the accurate identification of local values ​​of the single scattering rate.

[0160] The method of this disclosure then enables the identification of a decreasing function F within the liver. This function exhibits a representation of the mean free path of scattering within the liver. s The decay behavior. This is determined using theoretical curves. By fitting experimental points, this method thus enables the clear differentiation of two regions by identifying local mean free path within the liver: s~20mm. Practitioners can then use these differences clinically to provide advice or make a diagnosis.

[0161] An ultrasound characterization system according to the present disclosure for medical analysis of a medium M is shown in Figure 3. It includes: - a transducer array 10 configured to generate a series of incident ultrasound waves within a region of the medium and to measure the backscattered ultrasound waves from said region as a function of time; and - a processing unit (30) connected to the transducer array and configured to perform a method including the following steps: - generating a series of incident ultrasound waves US S110 within a region of the medium via the array 10 of transducers 11. in The series of incident ultrasonic waves form a transmitting base i; and - the canonical reflection matrix R is defined between the transmitting base i at the input and the receiving base (u) at the output of S120. ui (t), where the coefficients of the canonical reflection matrix correspond to the signals generated by the ultrasonic waves received by the transducer and reflected within the medium; - by using the canonical reflection matrix R ui (t) Focus on the spatial location r in = (x) in The virtual input transducer (z) with spatial position r out = (x) out Between the virtual output transducers of z), determine the focusing reflection matrix R of the S130 medium. xx (z), the two virtual transducers are located at the same depth z for the sound velocity model c0, and the focusing reflection matrix R xx The coefficients of (z) are written as follows: R xx (z) = [R( )],x in , x out Included around the point Within the area, virtual transducer r in and r out All lateral positions x in and x out A focusing basis (x) is formed at each depth z.

[0162] The method is characterized in that it further includes the following steps: - Focusing the reflection matrix Filtering S140 is performed to obtain the single-scattering matrix representing the single-scattering component of the focused reflection matrix. Or, representing the multiple reflection matrix of the multiple scattering components of the focused reflection matrix. - By calculating the single reflection matrix R s The norm of (z) and the focused reflection matrix R xx The ratio of the norms of (z) determines the S150 around the point. The single scattering rate as a function of depth z within the region Alternatively, by calculating the reflection matrix R multiple times. M The norm of (z) and the focused reflection matrix R xx The ratio of the norms of (z) determines the area around the point. The multiple scattering rate as a function of depth z within the region .

Claims

1. An ultrasound characterization method (S100) for media used in medical analysis, comprising the following steps: - A series of incident ultrasonic waves (US) are generated (S110) within the region of the medium by an array (10) of transducers (11). in The series of incident ultrasonic waves form a transmitting base (i); and the canonical reflection matrix R is defined by measuring (S120) between the transmitting base (i) at the input and the receiving base (u) at the output. ui (t), where the coefficients of the canonical reflection matrix correspond to the signals generated by the ultrasonic waves received by the transducer and reflected within the medium; - by using the canonical reflection matrix R ui (t) Focus on the spatial location r in = (x in The virtual input transducer (z) with spatial position r out = (x out Between the virtual output transducers of z), determine the focusing reflection matrix R of the (S130) medium. xx (z), the two virtual transducers are located at the same depth z for the sound velocity model c0, and the focusing reflection matrix R xx The coefficients of (z) are written as follows: R xx (z) = [R( )]x in , x out Included around the point Within the area, virtual transducer r in and r out All lateral positions x in and x out The method, which forms a focusing base (x) at each depth z, is characterized by further comprising the following steps: - For the focusing reflection matrix Filtering (S140) is performed to obtain the single reflection matrix representing the single scattering component of the focused reflection matrix. Or, representing the multiple reflection matrix of the multiple scattering components of the focused reflection matrix. - By calculating the single reflection matrix R s The norm of (z) and the focused reflection matrix R xx The ratio of the norms of (z) determines the (S150) around the point. The single scattering rate as a function of depth z within the region Alternatively, by calculating the reflection matrix R multiple times. M The norm of (z) and the focused reflection matrix R xx The ratio of the norms of (z) determines the area around the point. The multiple scattering rate as a function of depth z within the region 。 2. The method according to claim 1, characterized in that, Filtering (S140) is performed by: - ​​Determining (S141) a set of numerical focusing reflection matrices representing single scattering in the medium. It is achieved by using a transducer array with a spatial position in a model medium having a sound velocity c0. The acoustic propagation between point scatterers (x, z) is calculated. in , x out , , and z are included around the point Within the specified region, the coefficients of the numerically focused reflection matrix are written as follows: - Determine (S142) a confocal base, which includes a reflection matrix that is focused by at least a portion of the numerical values. Several confocal matrices calculated by orthogonal normalization For depth z, k is from 1 to N k The index between, N k To represent depth A positive integer representing the number of matrices in the confocal basis; - will focus the reflection matrix. Projection (S143) to the confocal matrix On the confocal basis, to obtain the single reflection matrix representing the single scattering component of the focused reflection matrix. Or, representing the multiple reflection matrix of the multiple scattering components of the focused reflection matrix. 。 3. The method according to claim 2, characterized in that: During the determination of the confocal base (S142), in the form of Numerical reflection matrix expressed Arranged as a two-dimensional matrix defined for each depth z In the form of a two-dimensional matrix Perform singular value decomposition according to the following expression: in Two-dimensional matrix singular values, Two-dimensional matrix Singular vectors in the focusing basis (x), Two-dimensional matrix In having spatial location ( The singular vectors in the point scattering volume basis of z, with superscripts This represents the combined operation of conjugation and transpose.

4. The method according to claim 2, characterized in that: For projection (S143), the single reflection matrix Calculated as a confocal matrix in a confocal basis With the focused reflection matrix A linear combination of inner product weights at each depth The projection is obtained using the following formula: in Representation matrix and The inner product between, and the superscript This indicates a conjugate operation.

5. The method according to claim 2, characterized in that: For projection (S143), the single reflection matrix Calculated as a confocal matrix in a confocal basis With the focused reflection matrix A linear combination of inner product weights at each depth The projection is obtained using the following formula: in Representation matrix and The inner product between, and the superscript This indicates a conjugate operation.

6. The method according to any one of claims 1 to 5, further comprising the following step: - By identifying single scattering rate Attenuation as a function of depth z, or by identifying the multiple scattering rate As the depth z increases, (S160) around the point is determined. The mean free path of scattering within the region s .

7. The method according to claim 6, further comprising: - For a set of points within the medium Determine the mean free path of scattering The image.

8. The method according to any one of claims 1 to 7, further comprising: - From the focused reflection matrix R xx The diagonal coefficients of (z) determine the confocal image I as a function of depth z. c (x,z), that is: 。 9. The method according to claim 8, further comprising: - Determine the surrounding point The average confocal intensity is obtained by averaging the confocal image along its lateral dimension x within the region. ： Among the symbols This represents the average over the index variable.

10. The method of claim 9, further comprising: - By identifying average confocal intensity The decay as a function of depth z determines (S170) around the point. Extinction length l within the region ext .

11. The method of claim 10, further comprising: - For a set of points within the medium Determine the extinction length The image.

12. The method according to claims 6 and 10, further comprising: - Determine the area around point (S180) using the following method. Absorption length l within the region a : 。 13. The method of claim 12, further comprising: - For a set of points within the medium Determine the absorption length The image.

14. An ultrasound characterization system (1) for a medium (M) used in medical analysis, the system comprising: - An array (10) of transducers (11) configured to generate a series of incident ultrasonic waves in a region of the medium and to measure the ultrasonic backscatter wave from said region as a function of time; and - a processing unit (30) connected to the transducer array and configured to implement the method according to any one of claims 1 to 13.

Images