WO2014113442A1

WO2014113442A1 - Systems and methods for estimating acoustic attenuation in a tissue

Info

Publication number: WO2014113442A1
Application number: PCT/US2014/011631
Authority: WO
Inventors: Jiangang Chen; Elisa E. Konofagou; Gary Yi HOU
Original assignee: The Trustees Of Columbia University In The City Of New York
Priority date: 2013-01-17
Filing date: 2014-01-15
Publication date: 2014-07-24

Abstract

Systems and techniques for estimating acoustic attenuation in a tissue from time-varying radiation force information generated through the application of acoustic energy to the tissue from at least first and second focal depths are provided. An exemplary technique includes acquiring first signals representing oscillatory motion of the tissue in response to the radiation force proximate the first focal depth, and acquiring second signals representing oscillatory motion of the tissue in response to the radiation force proximate the second focal depth. The technique further includes estimating the oscillatory motion of the tissue from each of the first and second signals, and estimating the acoustic attenuation in the tissue from the estimated oscillatory motion of the tissue from the first and second signals.

Description

SYSTEMS AND METHODS FOR ESTIMATING ACOUSTIC ATTENUATION IN

A TISSUE

CROSS REFERENCE TO RELATED APPLICATIONS

This application claims priority to U.S. Provisional Application No. 61/753,706, filed on January 17, 2013, which is incorporated by reference herein in its entirety.

STATEMENT REGARDING FEDERALLY- SPONSORED RESEARCH

This invention was made with government support from the National Institutes of Health under Grant No. R01 EBO 14496. The government has certain rights in the invention.

BACKGROUND

Acoustic attenuation generally refers to the reduction in acoustic pressure amplitude during propagation within a medium. The ability to accurately estimate attenuation can be useful in therapeutic ultrasound, where the acoustic intensity within the region of interest (ROI) can be estimated. This can allow for improved tracking of the induced temperature rise during tissue ablation; ultrasound imaging, where precise time gain compensation can be obtained to improve the image quality; and ultrasonic tissue characterization, which can allow for microscopic examination of the scatterer size and backscatter coefficient as well as in situ temperature monitoring. In the field of acoustic radiation force imaging, attenuation can be a factor for quantifying the generated radiation force. Furthermore, attenuation can be related to tissue pathology. For example, attenuation can be varied by a factor of up to 35% between normal and alcoholic livers in human subjects, which can provide an indicator for alcoholic liver disease. In addition, attenuation can correlates with pathologic fat and fibrosis in livers. Tissue attenuation can also change during lesion formation using HIFU (high intensity focused ultrasound).

One technique for estimating acoustic attenuation is the broadband substitution method. Other techniques can include centroid and multi-narrowband techniques, which can analyze backscattered ultrasound signals in B-mode images. Applications of such techniques can include estimating the differential attenuation of HIFU-induced lesions. Although certain techniques can be suitable for estimating tissue attenuation, the application of such techniques in clinical practice can be challenging, due at least part to the diffraction effect from the finite aperture of the transducer, which can introduce undesired spectral disturbance to the acoustic wave; influence of overlying tissues, for example the abdominal wall structure, which can distort the acoustic wave in spectrum due at least in part to phase aberration effects; and effects from scatterers, which can also influence the spectrum of the

backscattered signal.

Certain acoustic radiation force techniques can be utilized for attenuation measurements. For example, the reduction in radiation force resulting from the insertion of a tissue sample between a transducer and a reflector can be measured for attenuation estimation. Furthermore, an attenuation estimation approach using linear array transducers can be utilized to generate a radiation force. The induced displacement can be monitored after the application of the radiation force. The ultrasound focus can be electronically shifted away from the transducer surface while keeping the f-number of the transducer constant, and the attenuation can be calculated at the focal depth, which can be where the radiation force reaches a maximum. Such techniques can be applied using conventional diagnostic scanners without additional hardware.

Harmonic Motion Imaging (HMI) is another example of a radiation- force-based technique. However, HMI can include monitoring the displacement in synchronization with the application of radiation force, which can provide tissue properties that certain other techniques cannot. HMI can also be used to monitor thermal ablation based on the displacement variations due to changes in tissue stiffness during ablation, and to evaluate changes in the tissue viscoelasticity parameters. Improving the ability of HMI to quantify the Young's modulus of soft tissues can be beneficial in implementing clinically translatable mechanical testing systems and techniques for in vivo application. However, the radiation force exerted within the excitation region is not necessarily known.

SUMMARY

Techniques for estimating acoustic attenuation in a tissue are disclosed herein. In one embodiment of the disclosed subject matter, methods are provided for estimating acoustic attenuation in a tissue from time-varying radiation force information generated through the application of acoustic energy to the tissue from at least first and second focal depths. An example method includes acquiring first signals representing oscillatory motion of the tissue in response to the radiation force proximate the first focal depth, and acquiring second signals representing oscillatory motion of the tissue in response to the radiation force proximate the second focal depth. The method further includes estimating the oscillatory motion of the tissue from each of the first and second signals, and estimating the acoustic attenuation in the tissue from the estimated oscillatory motion of the tissue from the first and second signals.

In some embodiments, the method can include applying the acoustic energy by pulsing a focused ultrasound transducer at a modulation frequency.

Acquiring each of the first and signals can include pulsing an imaging transducer configured as a pulser/receiver to acquire radio frequency signals at a pulse repetition frequency.

In some embodiments, estimating the oscillatory motion of the tissue from each of the first and second signals can include applying ID normalized cross correlation to the acquired radio frequency signals. Additionally or alternatively, estimating the acoustic attenuation can include linearly correlating the estimated oscillatory motion from each of the first and second signals.

In some embodiments, the method can include estimating the acoustic attenuation at a first portion of the tissue, estimating the acoustic attenuation at a second portion of the tissue lateral from the first portion, and determining a displacement map of the tissue using the estimated acoustic attenuation of the first portion and the estimated acoustic attenuation of the second portion.

In another aspect of the disclosed subject matter, systems for estimating acoustic attenuation in a tissue are provided, and generally include an ultrasound transducer an imaging transducer, one or more memories and a processor. In an example, the ultrasound transducer is configured to apply acoustic energy to the tissue a first focal depth and a second focal depth to generate a time-varying radiation force proximate the first focal depth and the second focal depth. The imaging transducer is configured to be optically coupled to the tissue and acquire first signals representing oscillatory motion of the tissue in response to the radiation force proximate the first focal depth and second signals representing oscillatory motion of the tissue in response to the radiation force proximate the second focal depth. The one or more processors are coupled to the one or more memories and the imaging transducer and configured to estimate the oscillatory motion of the tissue from each of the first and second signals; and estimate the acoustic attenuation in the tissue from the estimated oscillatory motion of the tissue from the first and second signals.

In some embodiments, the one or more processors can be coupled to the ultrasound transducer and can be further configured to pulse the ultrasound transducer at a modulation frequency. The imaging transducer can be configured as a pulser/receiver, and in some embodiments, the one or more processors can be further configured to pulse the imaging transducer at a pulse repetition frequency to acquire radio frequency signals corresponding to each of the first and second signals.

In some embodiments, the one or more processors can be further configured to estimate the oscillatory motion of the tissue from each of the first and second signals by applying ID normalized cross correlation to the acquired radio frequency signals. Estimating the acoustic attenuation can include linearly correlating the estimated oscillatory motion from each of the first and second signals.

In some embodiments, the system can include a positioning apparatus coupled to the ultrasound transducer and logically coupled to the one or more processors. The positioning apparatus can be to move the ultrasound transducer to aim the ultrasound transducer at the first focal depth and the second focal depth in response to the one or more processors. The positioning apparatus can be further configured to aim the ultrasound transducer to a first portion of the tissue and a second portion of the tissue lateral from the first portion in response to the one or more processors. As such, the one or more processors can be further configured to estimate the acoustic attenuation at each of the first portion and second portion of the tissue, and determine a displacement map of the tissue using the estimated acoustic attenuation of the first portion and the estimated acoustic attenuation of the second portion. The imaging transducer can be coupled to and coaxially aligned with the ultrasound transducer.

BRIEF DESCRIPTION OF THE DRAWINGS

The accompanying drawings, which are incorporated and constitute part of this disclosure, illustrate some embodiments of the disclosed subject matter. FIG. 1 is a diagram illustrating an exemplary system for estimating acoustic attenuation according to the disclosed subject matter.

FIGS. 2(a)-2(b) are diagrams illustrating attenuation measurements at exemplary focal locations in a tissue.

FIGS. 3(a)-3(b) are images of exemplary HIFU lesions in in vitro canine livers.

FIG. 4 is a diagram illustrating exemplary harmonic variation in local displacement.

FIGS. 5(a)-5(j) are diagrams illustrating HMI displacement compared to acoustic intensity.

FIGS. 6(a)-6(e) are exemplary 2D HMI displacement maps.

FIG. 7 is a diagram illustrating normalized HMI displacement compared to depth in exemplary phantoms.

FIGS. 8(a)-8(e) are diagrams illustrating linear regression for determining attenuation in exemplary phantoms.

FIGS. 9(a)-9(c) are diagrams illustrating estimated attenuation compared to (a) independent measurement of exemplary phantoms, (b) Bland-Altman analyses of the measurements of (a), and (c) estimation errors of estimated attenuations with respect to independent measurement.

FIGS. 10(al)-(b2) are diagrams illustrating HMI displacement depth and linear regression under different acoustic intensities.

FIGS. 1 l(al)-(c3) are (al)-(a3) HMI displacement maps and (bl)-(b3) HMI displacement curves obtained from three in vitro canine livers, and (cl)-(c3) linear regression for estimating attenuation of the livers.

FIG. 12 is a diagram illustrating estimated attenuations of liver tissues before and after ablation using different acoustic powers.

FIGS. 13A-13C illustrate exemplary HMI displacement maps, before HIFU exposure, after HIFU exposure, and after minus before (i.e., HMI displacement contrast map), respectively.

Throughout the figures and specification the same reference numerals are used to indicate similar features and/or structures. DETAILED DESCRIPTION

The systems and methods described herein can be useful for estimating acoustic attenuation from time-varying radiation force information generated through the application of acoustic energy. Although the description provides as an example estimating acoustic attenuation of a biological system, such as biological tissue, the systems and methods herein can be useful for estimating acoustic attenuation of any suitable system that provides radiation force information through the application of acoustic energy.

The subject matter disclosed herein includes methods and systems for estimating acoustic attenuation in a tissue. Accordingly, they can utilize time-varying radiation force information generated through the application of acoustic energy to the tissue from at least first and second focal depths. An exemplary technique includes acquiring first signals representing oscillatory motion of the tissue in response to the radiation force proximate the first focal depth, and acquiring second signals representing oscillatory motion of the tissue in response to the radiation force proximate the second focal depth. The method further includes estimating the oscillatory motion of the tissue from each of the first and second signals, and estimating the acoustic attenuation in the tissue from the estimated oscillatory motion of the tissue from the first and second signals

In accordance with the disclosed subject matter, estimating acoustic attenuation in a tissue can include estimating acoustic attenuation of biological tissues using HMI with a linear regression model. HMI can provide oscillatory information from displacement induced in a tissue, and resulting harmonics can be separated from quasi-static effects. Furthermore, HMI can provide a localized result at least in part because the displacement can be measured at the focus of the FUS transducer. In this manner, attenuation estimation using HMI can provide a quantitative technique for both elasticity imaging of soft tissue and assessment of tissue elasticity undergoing thermal ablation such as HIFU.

With reference to FIG. 1, an exemplary HMI system 100 according to the disclosed subject matter can generally include an action unit 102 and a control unit 104. The action unit 102 can include a focused ultrasound (FUS) transducer 106, for example and without limitation, embodied herein as a PZT transducer configured with a focal depth of 90 mm and center frequency ( f_center = 4.75 MHz ), and an imaging transducer 108, for example and without limitation, a concentric and confocal single-element, pulse-echo transducer (embodied herein as a V30044 Immersion Transducer, focal depth: 6 cm, f_center = 7.5 MHz , Olympus-NDT,

Waltham, MA, USA). FUS transducer 106 and imaging transducer 108 can be confocally aligned and attached to a 3D positioning system 1 10. The control unit 104 can include a processor 1 12 operably coupled to the action unit 102. For example and without limitation, the processor 112 can be embodied as a PC workstation (CPU: 3.06 GHz; RAM: 80 GB), and can be operably coupled to the action unit by one or more control lines. For purpose of illustration and not limitation, three control lines are utilized from the processor 110, one for each of the FUS transducer 106 (also referred to as "Control line 1"), the imaging transducer 108 (also referred to as "Control line 2"), and the 3D positioning system 1 10 (also referred to as "Control line 3").

For purpose of illustration, and as embodied herein, processor 1 12 can output an amplitude-modulated (AM) signal, for example and without limitation at a carrier frequency of 4.75 MHz, via Control Line 1 using a first signal generator 1 14 (also referred to as Function Generator 1, embodied herein as Model: 33220A, Agilent^®, CA, US) and a modulation frequency, for example and without limitation at 25 Hz, via Control Line 1 using a second signal generator 1 16 (also referred to as Function Generator 2, embodied herein as Model: 33120A, HP^®, NY, US). The activation duration of each signal can be 400 ms, and a duration between two adjacent bursts can be 1 s. The AM signal from signal generator 114 can be amplified through a PvF power amplifier (for example, embodied here as Model: 3000L, ENI^®, NY, USA), and thus can have an acoustic intensity of 0.1 W/cm² on the transducer surface. The acoustic intensity can be obtained by dividing the acoustic power (for example as measured using a radiation force balance) by the area of the active surface of the focused transducer (as embodied herein, the area of the transducer active surface = 50.87 cm ). In this manner, the AM ultrasonic wave can induce a time-varying radiation force in the focal region of the FUS transducer, which can occur at twice the modulation frequency (i.e., 50 Hz). Oscillatory motion can thus occur at the focal zone, and can be detected by the imaging transducer 108 during force application.

Furthermore, and as embodied herein, the processor 1 10 can operate the imaging transducer 108 via Control line 2, for example in a pulse-echo manner using a pulser/receiver 1 18 (embodied herein as 5800PR, Olympus NDT , NY, USA) for acquiring RF signals at a pulse repetition frequency (PRF), for example and without limitation at 4 kHz, and can occur in conjunction with the operation of Control line 1. The captured RF signals can be input into a band-pass filter for filtering out the carrier frequency, and can be digitized by a data acquisition board (Gage^®, IL, USA), embodied herein with a sampling frequency of 100 MHz. ID normalized cross correlation can be applied to the RF signals for estimating the oscillatory motion, embodied herein with a window size of 1 mm and 90% overlap.

In HMI according to the disclosed subject matter, the acoustic energy emitted by the FUS transducer 102 can converge at the transducer focus, and thus a radiation force can be locally generated, the magnitude of which can be represented as

^ =— , (1) c

where c (e.g., 1540 mis), f , a = a(f) (dB/cm) and / (W/cm²) can represent the sound speed, frequency, frequency-dependent attenuation coefficient of the tissue and in situ temporal average intensity, respectively. The intensity (/ ) can be determined from the acoustic pressure ( p ) according to

< = , (2)

2pc

where can represent the density of the medium. The radiation force can be obtained from the acoustic pressure according to

F . (3) pc

The activation surface of the FUS transducer can be represented as a concave spherical geometry modeled as a set of equivalent monopole sources uniformly distributed over the transducer aperture and excited in phase, and thus the pressure distribution of such a radiator can be approximated by Rayleigh function, for example in the form of an integral taken over the area of the transducer surface in a non-attenuating medium. The attenuation and dispersion effects associated with the transmission of the ultrasonic beam in an attenuating material can be represented as the complex wavenumber k_c, for example as where z = V-T. As such, the pressure field at the focus in an attenuating homogeneous medium can be determined from e . (4) as

where R , p₀ , λ and a can represent the focal radius, acoustic pressure at the transducer surface, wavelength and transduce radius, respectively.

In the exemplary system and technique described herein, the wave propagation path can be considered to cover biphasic media: that is, water and tissue {i.e., an inhomogeneous medium). As such, the result of eq. (5) can be determined through the definition of a single medium using an effective attenuation coefficient without nonlinearity, which can be represented as

where z_w and z_t can represent the propagation depths of the beam within the water and tissue, respectively. Furthermore, the attenuation of water can be relatively negligible.

Two different focal positions in the tissue can be represented with depth being respectively z_n (as illustrated in FIG. 2(a)) and z_l2 (as illustrated in FIG.

2(b)), and a reflection at the water-tissue interface can occur. As such, the ratio between the acoustic pressures at the foci of the two positions, as illustrated in FIGS. 2(a)-2(b), can be represented as

where t t₂ can represent the transmission coefficients between water and tissue in FIG. 2(a) and 2(b), respectively, and a_iffl (.f) and a_eff2 (f) can represent the effective attenuations in FIG. 2(a) and 2(b), respectively.

At the focus of the FUS transducer, for example where ^z _wX + ^z = ^R (FIG. 2(a)) and ^z _w2 ^{+ z} _l2 = ^R (FIG. 2(b)), eq. (7) can be represented as P_\{R) ₌ ^a -^» ₌ f, -_g, (,„-.„) ₍₈

p₂(R) t₂e^~a^ t₂

As discussed herein, can represent the frequency-dependent attenuation coefficient of the tissue, i.e., a = a(f) . The attenuation of the soft tissue can be linearly correlated with frequency as a first-order approximation, and thus a_t(f) = a_tf in eq. (8). The transmission coefficients t_x and t₂ can be considered as identical, and as such, the two media (i.e. , water and tissue) can remain the same in both cases, and the wave incident angle can change only insignificantly when z and z_a are disposed a small distance apart, for example and without limitation, embodied herein as 5 mm. Furthermore, the radiation force (F ) can change linearly based at least in part on the square of the acoustic pressure (eq. (3)). As such, the ratio between the radiation forces at depths z„ and z_t2 can be expressed by

^F\ (^R) _ _e-2«,/(-- i--,₂) m

F₂(R)

The attenuation coefficient can thus be obtained, for example by

As such, the acoustic attenuation can be obtained from the ratio between F(R) and

F₂(R) , frequency ( / ), and distance between z_n and z_a . The radiation force at the focus of the FUS transducer (i.e., F_{(R) and F₂(R) ) can be obtained from testing samples by varying the output intensity of the FUS transducer (embodied herein as 0.03—0.22 W/cm²), and represented as I = k'D_{; wnere} £>can represent HMI displacement, / can represent output intensity of the FUS transducer and k' can represent a linear coefficient, as discussed further herein. Such an examination can be performed at multiple, pre-selected focal positions covering the whole raster-scan plane in the sample, for example and embodied herein using 5 positions at each focal depth. Furthermore, the intensity and the induced radiation force can be linearly proportional at certain frequencies, for example the HMI carrier frequency utilized herein, and thus can be represented as F = k"I^ _where ^ ^can represent output intensity of the FUS transducer, and k" linear coefficient, which can be determined from equation (1). The ratio between the radiation forces can thus be represented as equal to that of displacements, that is

-^- =—, (11)

F₂ (R) D₂

where Z , and D₂ can represent displacements induced by F_l and F₂ at depths z and z_l2 , respectively. As such, HMI-related acoustic attenuation can be represented as

The technique for obtaining the representation of eq. (12) can be applied to the HMI displacements estimated at different depths for attenuation estimation, and thus the displacement at every depth ( D_z ) can be compared with that at the initial depth (

D₀ ), which can be represented as

where z₀ , z , D₀ and D_z can represent the initial depth, the arbitrary depth, the HMI displacements at the initial depth and arbitrary depth, respectively. In this manner, the attenuation can be estimated using a linear regression model, for example and as ln(^> )

D

embodied herein, by linearly correlating — and (z - z₀) . Such a technique can

2f

thus utilize differences in HMI displacement at different focal depths, where attenuation effect can be a factor in the decrease in acoustic energy when the focus deepens.

The techniques described herein can be applied, for purpose of illustration and confirmation of the disclosed subject matter, and not limitation, to estimate attenuation in five phantoms with known attenuations (Computerized Imaging Reference Systems (CIRS), Inc., VA, USA) (as shown in Table 1). The phantoms can include three normal canine livers in vitro and five canine livers in vitro after HIFU ablation. The phantoms, for illustration and not limitation, and as embodied herein, can have dimensions of 50 mm in diameter and 50 mm in height, and can have homogeneous material properties. The attenuation of each phantom can be measured using log spectral difference measurement, with the parameters listed in Table 1. Each phantom can be immersed in degassed water in a water tank during the measurement with the phantom sealed using a thin membrane to avoid water ingress. Rubber absorbers can be placed between the phantom and edges of the water tank to avoid reflections of the ultrasound waves, as illustrated for example in FIG. 1. The phantoms can be tested using HMI, as described herein, with the confocally-aligned FUS transducer 106 and imaging transducer 108 operated in a raster-scan format, for example at a scanning step of 0.5 mm with the total scanned area of 5><5 mm² in the y-z plane. The scanned region can be chosen to be at least 3 mm deep from the upper surface of the phantom to avoid any boundary effects.

Furthermore, for purpose of illustration and confirmation of the disclosed subject matter, to determine the effect of the output acoustic intensity of the FUS transducer on the results, two additional examples can be conducted (herein on Phantom 2), embodied herein with (i) the acoustic intensity at 0.1 W/cm² (as performed above) and the raster scanning depth of 10 mm at a scanning step of 1 mm; and (ii) with the same configuration described herein but with the acoustic intensity increased to 0.2 W/cm². As embodied herein, Phantom 2 can have an attenuation of 0.57 dB/cm/MHz, which can represent an attenuation of biological tissues.

Additionally, for purpose of illustration and confirmation of the disclosed subject matter, the techniques described herein were applied to estimate the attenuations of three in vitro canine livers obtained from three mongrel male dogs. Each specimen was immersed in phosphate buffered saline (PBS) solution and placed in a vacuum chamber for one and a half hours for degassing. The liver tissues were moved from the vacuum chamber to the water bath filled with degassed PBS solution, and the samples remained submerged in degassed saline to avoid air exposure. The attenuation measurement in the liver remained the same to that of phantoms.

In addition, for purpose of illustration and confirmation of the disclosed subject matter, five in vitro canine livers were ablated using the FUS transducer 102 excited (i) at 600 mV (using Function generator 1, providing an acoustic intensity of about 0.1 W/cm² at the surface of the FUS transducer) with an activation duration of 120 s, or (ii) at 900 mV (using Function generator 1, providing an acoustic power of around 0.21 W/cm² at the surface of the FUS transducer) for 30 s. In these examples, the FUS transducer was operated in a raster-scan manner, i.e., with 1 1 consecutive positions with the moving step of 3 mm in the lateral direction and 2 positions offset by 3 mm in the axial direction, providing a lesion with the dimension of roughly 2x3 cm², as illustrated in FIG. 3(a) for acoustic intensity of 0.1 W/cm² and FIG. 3(b) for acoustic intensity of 0.21 W/cm². The attenuation of each liver was measured before ablation, to provide a reference, and two different acoustic powers were chosen to determine the effect of different acoustic intensities during HIFU ablation on the attenuation of the induced lesions.

Example

As a representative example, FIG. 4 illustrates the displacement curve at the focus of the FUS transducer over 100 ms, embodied herein with an output intensity at the transducer surface of 0.1 W/cm², captured from Phantom 1. As discussed herein, the linear elasticity of each sample was evaluated by varying the output acoustic intensity of the FUS transducer. The relationship between the acoustic intensity and induced displacement in the five phantoms, three in vitro normal livers and five livers with HIFU lesions is illustrated in FIG. 5 (FIGS, 5(a)- 5(e) for phantoms 1-5, FIGS. 5(f)-5(h) for three liver samples and FIGS. 5(i)-5(j) for HIFU lesions produced using two different acoustic intensities, respectively), with the correlation coefficient varying within 0.814-0.982.

The HMI displacements were estimated in all raster-scan locations, forming a 2D HMI displacement map, as shown for example in FIG. 6 (for phantoms 1-5). The average HMI displacement, which can correspond to the average peak-to- peak HMI displacements over the duration of the HMI application, as shown for example in FIG. 4, was obtained at different depths and each compared in FIG. 7. FIGS. 8(a)-8(e) illustrate the estimated attenuations of phantoms 1-5 using the linear regression model (i.e. , equation (13)), respectively, which are also listed in Table 1. The estimated attenuations were compared with those independently measured (from Table 1), with the correlation coefficient equal to 0.976 (as illustrated in FIG. 9(a)). FIG. 9(b) illustrates a Bland-Altman analysis of the data, and FIG. 9(c) illustrates estimation errors that varied within 15%-35%. FIG. 10 shows the estimated displacements and attenuations using different acoustic intensities (i.e., 0.1 and 0.2 W/cm²) to illustrate effects of the output acoustic intensity of the FUS transducer on the techniques of the disclosed subject matter. The attenuation in three in vitro canine livers, as shown for example in Table 2, varied in a range from 0.293 to 0.353 dB/cm/MHz. For purpose of illustration, FIG. 1 1 presents the displacement maps (FIGS. I l(al)-l l(a3)) and plots (FIGS. I l(bl)-l l(b3)) obtained from three in vitro canine livers, from which the estimated attenuations are illustrated in FIGS. 1 l(cl)-l l(c3), respectively.

The estimated attenuations in in vitro canine livers before and after ablation, i.e., HIFU lesions, are illustrated, for purpose of comparison, in FIG. 12 and Table 3. Paired-sample t-test evaluation showed differences in estimated attenuation between the normal tissue and HIFU lesions using different acoustic intensities, that is, p-values of 0.0018 for the tissue before and after ablation at acoustic intensity of 0.1 W/cm², 1.06* 10^"4 for the tissue before and after ablation at acoustic intensity of 0.21 W/cm² and 0.0383 for HIFU lesions using different acoustic intensities. As such, the HIFU lesions were measured to have higher attenuation than the normal tissues, and of the HIFU lesions, those produced using higher acoustic intensity were estimated to have higher attenuation.

Attenuation estimation using HMI, according to the disclosed subject matter, can be performed to simultaneously generate the radiation force and monitor the induced local displacement at the focus of the FUS transducer. The HMI displacements estimated at different depths within the raster-scan plane can be analyzed using a linear regression model for estimating the attenuation. In this manner, the local displacement (i.e. , HMI displacement) can vary with depth, and thus localized tissue attenuation within a region can be estimated. As such, the techniques according to the disclosed subject matter can be applied to evaluate tissues with regional inhomogeneity, for example and without limitation, tumors and HIFU- induced thermal lesions. Furthermore, as discussed herein, the techniques can be performed independent of the speed of sound, as shown for example in eqs. (3) and (9), and as such, tissues can be evaluated under thermal treatment.

In the examples discussed herein, the HMI displacements were estimated at all the raster-scan locations to form the 2D HMI displacement maps, as shown for example in FIG. 6. In the phantoms, in vitro livers and HIFU lesions of the examples, a linear relationship between the acoustic intensity and HMI displacement was determined, as illustrated for example in FIG. 5. Such a relationship was determined for all sample points selected throughout the raster-scan plane. As such, the samples were considered as linearly elastic at the force amplitude and frequency used. In addition, soft tissues can generally be considered as linearly elastic at low excitation frequencies (e.g., less than 100 kHz), and the tissue harmonic motion induced by the radiation force, as embodied herein, was determined to be 50 Hz, which further supports the linear relationship of the displacements.

Referring now to FIG. 6, the HMI displacement decreased with depth, due at least in part to the effect of attenuation. The decreasing rates in HMI displacement were different for phantoms of different attenuations, that is, for phantoms of higher attenuations, the HMI displacements decrease at a higher rate, as illustrated for example in FIG. 7. As such, the attenuations of those phantoms were differentiated using the linear regression model of eq. (13), as shown for example in FIG. 8 having high correlation coefficients. With reference to FIG. 8, the correlation coefficient for Phantom 5 (0.863) with the higher attenuation is lower than those for Phantoms 1-4 (about 0.98). The correlation coefficients can be due, at least in part, to an HMI displacement decreasing rapidly with depth for the phantoms with relatively high attenuation, as illustrated in FIG. 7. As a result, the signal-to-noise ratio (SNR) of the technique decreased.

In human soft tissues, attenuation can relate to various pathological conditions, as discussed above. The attenuations estimated using the proposed technique were found to linearly correlate with those independently measured, as shown for example in FIG. 9(a), with a linear correlation coefficient of 0.976, and analyzed by Bland-Altman analysis, as shown for example in FIG. 9(b). The estimation errors are illustrated in FIG. 9(c), and vary within the range of about 15%- 35%. Furthermore, the phantom with the highest attenuation (e.g., Phantom 5) exhibited the largest estimation error. The estimation error can be due, at least in part, to the displacement in the high attenuated phantom decreasing quickly with depth. Furthermore, attenuated displacements beyond certain depths can introduce noise at least in part due to resolution limits of the equipment, and as such, the noise can deteriorate the estimated value and introduce estimation errors and reduce SNR. The estimation errors are relatively smaller for phantoms of low attenuations (e.g., Phantoms 1 and 2), which can represent superficial tissue regions or other tissues with lower attenuations, e.g., in vivo livers. Additionally, SNR can be increased by raising the acoustic intensity, which can result in higher HMI displacement in higher attenuating materials. With reference to FIG. 10, the estimation using the higher acoustic intensity has a higher regression coefficient (0.985), and the attenuation value (0.436 dB/cm/MHz) is closer to that independently measured (0.57 dB/cm/MHz in Table 1) than the estimate using the lower acoustic intensity. As such, the estimation accuracy and sensitivities in high attenuating materials can be improved by increasing the acoustic intensity.

Furthermore, the higher intensity (i.e., 0.2 W/cm²) remained within the range in which the phantoms can be tested and deemed to be linearly elastic, as shown for example in FIG. 5, and thus eq. (13) can be applied to the attenuation estimation. With reference to FIG. 10a(2), the attenuation obtained using the higher intensity (i.e., 0.436 dB/cm/MHz) was compared with that using the low intensity (i.e., 0.376 dB/cm/MHz) to illustrate the effect of the estimated attenuation on the acoustic intensity. The accuracy of the estimated attenuation at higher acoustic intensities can be affected at least in part by thermal effects under higher acoustic intensities. This can be due at least in part because the tissue stiffness can be changed, and thus the HMI displacements can be altered. Furthermore, the tissue attenuation can be temperature-dependent. As discussed herein, the HMI sequence can be configured to produce displacement measurement with suitable SNR while reducing or minimizing any thermal/nonlinear effects.

Referring now to FIGS. I l(al)-l l(a3) and 11 (bl)-l 1 (b3), HMI displacement maps and curves, respectively, from three in vitro canine livers are illustrated. A decreasing relationship of the HMI displacement with depth can be observed, similar to that of the phantoms in FIGS. 6 and 7. The corresponding linear regression results are illustrated in FIGS. I l(cl)-l l(c3), with the linear regression coefficient being around 0.95, which can indicate that the liver is relatively homogeneous. The estimated acoustic attenuation of the canine livers was determined to be 0.32±0.03 dB/cm/MHz (as shown in Table 2) using the techniques of the disclosed subject matter, which corresponds to suitable ranges for in vitro normal livers (e.g., 0.28-0.399 dB/cm/MHz).

The phantoms used in this study were of homogeneous material property, which can correspond to the high linear regression coefficients, as shown in FIG. 7. For the in vitro livers tested in this study, the homogeneity likewise corresponds to linear regression coefficients of around 0.95, as shown for example in FIGS. 9(cl)-9(c3). The HMI displacement can be considered a local measurement, and thus inhomogeneity can introduce undesired bias to the estimation results. Increasing the raster-scan density, e.g., reducing the scanning differential, can reduce the effect of such bias, and can be used to provide an attenuation map of inhomogeneous tissues.

Referring now to FIG. 12, the HIFU lesions had higher attenuation than the normal tissues. Furthermore, the HIFU lesion produced under the acoustic intensity of 0.21 W/cm² was shown to have higher attenuation compared to those under the intensity of 0.1 W/cm², due at least in part because the higher intensity can change the property of the tissue more severely and increase attenuation. As such, the techniques according to the disclosed subject matter can be performed on inhomogeneous tissues, e.g., HIFU lesions, as well as homogeneous tissues. Figures 13A-13C illustrate the exemplary qualitative attenuation maps before (FIG. 13 A) and after (FIG. 13B) HIFU ablation in a canine liver, as well as the difference between the before and after maps (FIG. 13C), which can highlight the location of lesions.

The foregoing merely illustrates the principles of the disclosed subject matter. Various modifications and alterations to the described embodiments will be apparent to those skilled in the art in view of the teachings herein. It will thus be appreciated that those skilled in the art will be able to devise numerous techniques which, although not explicitly described herein, embody the principles of the disclosed subject matter and are thus within its spirit and scope.

Claims

1. A computer-implemented method for estimating acoustic attenuation in a tissue from time-varying radiation force information generated through the application of acoustic energy to the tissue from at least first and second focal depths, comprising:

acquiring first signals representing oscillatory motion of the tissue in response to the radiation force proximate the first focal depth;

acquiring second signals representing oscillatory motion of the tissue in response to the radiation force proximate the second focal depth;

estimating, by the processor, the oscillatory motion of the tissue from each of the first and second signals; and

estimating, by the processor, the acoustic attenuation in the tissue from the estimated oscillatory motion of the tissue from the first and second signals.

2. The method of claim 1, further comprising applying the acoustic energy by pulsing a focused ultrasound transducer at a modulation frequency.

3. The method of claim 1, wherein acquiring each of the first and signals comprises pulsing an imaging transducer configured as a pulser/receiver to acquire radio frequency signals at a pulse repetition frequency.

4. The method of claim 3, wherein estimating the oscillatory motion of the tissue from each of the first and second signals comprises applying ID normalized cross correlation to the acquired radio frequency signals.

5. The method of claim 1, wherein estimating the acoustic attenuation comprises linearly correlating the estimated oscillatory motion from each of the first and second signals.

6. The method of claim 1 , further comprising estimating the acoustic attenuation at a first portion of the tissue, estimating the acoustic attenuation at a second portion of the tissue lateral from the first portion, and determining a displacement map of the tissue using the estimated acoustic attenuation of the first portion and the estimated acoustic attenuation of the second portion.

7. A system for estimating acoustic attenuation in a tissue, comprising: an ultrasound transducer configured to apply acoustic energy to the tissue a first focal depth and a second focal depth to generate a time-varying radiation force proximate the first focal depth and the second focal depth; an imaging transducer configured to be optically coupled to the tissue and acquire first signals representing oscillatory motion of the tissue in response to the radiation force proximate the first focal depth and second signals representing oscillatory motion of the tissue in response to the radiation force proximate the second focal depth;

one or more memories; and

one or more processors coupled to the one or more memories and the imaging transducer, wherein the one or more processors are configured to:

estimate the oscillatory motion of the tissue from each of the first and second signals; and

estimate the acoustic attenuation in the tissue from the estimated oscillatory motion of the tissue from the first and second signals.

8. The system of claim 7, wherein the one or more processors is coupled to the ultrasound transducer and further configured to pulse the ultrasound transducer at a modulation frequency.

9. The system of claim 7, wherein the imaging transducer is configured as a pulser/receiver, and the one or more processors is further configured to pulse the imaging transducer at a pulse repetition frequency to acquire radio frequency signals corresponding to each of the first and second signals.

10. The system of claim 9, wherein the one or more processors is further configured to estimate the oscillatory motion of the tissue from each of the first and second signals by applying ID normalized cross correlation to the acquired radio frequency signals.

11. The system of claim 7, wherein estimating the acoustic attenuation comprises linearly correlating the estimated oscillatory motion from each of the first and second signals.

12. The system of claim 7, further comprising a positioning apparatus coupled to the ultrasound transducer and logically coupled to the one or more processors, the positioning apparatus configured to move the ultrasound transducer to aim the ultrasound transducer at the first focal depth and the second focal depth in response to the one or more processors.

13. The system of claim 12, wherein the positioning apparatus is further configured to aim the ultrasound transducer to a first portion of the tissue and a second portion of the tissue lateral from the first portion in response to the one or more processors, the one or more processors further configured to:

estimate the acoustic attenuation at each of the first portion and second portion of the tissue, and

determine a displacement map of the tissue using the estimated acoustic attenuation of the first portion and the estimated acoustic attenuation of the second portion.

14. The system of claim 7, wherein the imaging transducer is coupled to and coaxially aligned with the ultrasound transducer.