NZ615910B2

NZ615910B2 - Vascular characterization using ultrasound imaging

Info

Publication number: NZ615910B2
Application number: NZ615910A
Authority: NZ
Inventors: Andrew J Casper; Emad S Ebbini; Dalong Liu; Yayun Wan
Original assignee: Regents Of The University Of Minnesota
Priority date: 2011-04-14
Filing date: 2012-04-13
Publication date: 2016-03-01

Abstract

imaging method and a system for vascular imaging are disclosed. The system comprises one or more ultrasound transducers and a processing apparatus. The one or more transducers are configured to deliver ultrasound energy to a vascular region resulting in pulse-echo data therefrom. The processing apparatus is configured to control the capture of pulse-echo data at a frame rate such that measured displacement of a vessel wall defining at least one portion of a blood vessel in the vascular region and measured average blood flow through the at least one portion of the blood vessel have a quasi-periodic profile over time to allow motion tracking of both the vessel wall and the blood flow simultaneously; generate strain and shear strain image data for the region in which the at least one portion of the vessel is located using speckle tracking. The speckle tracking comprises using multi-dimensional correlation of pulse-echo data of one or more speckle regions undergoing deformation in the region in which the at least one portion of a blood vessel is located and the multi-dimensional correlation comprises determining a cross-correlation peak for the sampled pulse-echo data based on phase and magnitude gradients ofthe cross-correlated pulse-echo data; and The processing apparatus is further configured to identify at least one vascular characteristic of the vascular region in which at least one portion of a blood vessel is located based on the strain and shear strain image data. The at least one vascular characteristic comprises at least one of a flow characteristic associated with flow through the blood vessel, a structural characteristic associated with the blood vessel, and a hemodynamic characteristic associated with the blood vessel g apparatus is configured to control the capture of pulse-echo data at a frame rate such that measured displacement of a vessel wall defining at least one portion of a blood vessel in the vascular region and measured average blood flow through the at least one portion of the blood vessel have a quasi-periodic profile over time to allow motion tracking of both the vessel wall and the blood flow simultaneously; generate strain and shear strain image data for the region in which the at least one portion of the vessel is located using speckle tracking. The speckle tracking comprises using multi-dimensional correlation of pulse-echo data of one or more speckle regions undergoing deformation in the region in which the at least one portion of a blood vessel is located and the multi-dimensional correlation comprises determining a cross-correlation peak for the sampled pulse-echo data based on phase and magnitude gradients ofthe cross-correlated pulse-echo data; and The processing apparatus is further configured to identify at least one vascular characteristic of the vascular region in which at least one portion of a blood vessel is located based on the strain and shear strain image data. The at least one vascular characteristic comprises at least one of a flow characteristic associated with flow through the blood vessel, a structural characteristic associated with the blood vessel, and a hemodynamic characteristic associated with the blood vessel

Description

PCT/U52012/033584 VASCULAR CHARACTERIZATION USING ULTRASOUND IMAGlNG ENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT {01] This invention was made with government support under award number EB 6893 from NIH. The government has n rights in this invention.

CROSS NCE TO RELATED APPLICATIONS This application claims the beneﬁt ofUS Provisional Application Serial No. 61/475,550, ﬁled 14 April 2011, entitled “Vascular Characterization Using ound Imaging,” which is incorporated herein by reference in its entirety.

BACKGROUND The disclosure herein relates generally to ultrasound imaging. More particularly, the disclosure herein pertains to ultrasound imaging methods and systems for use in, e.g., diagnostic and/or therapy ations (e.g., imaging ofblood vessels and/or s proximate thereto, etc). ar imaging is gaining sed attention not only as a way to detect cardiovascular diseases, but also for the evaluation of response to new anti- atherosclerotic therapies (see, Ainsworth, et 31., “3D ultrasound measurement of change in carotid plaque volume — A tool for rapid evaluation ofnew therapies,” Stroke, vol. 36, no. 9, pp. l904~1909, SEP 2005). Intravascular ultrasound (IVUS) has been shown to provide an effective tool in measuring the progression or regression ofatherosclerotic disease in response to therapies. However, IVUS is invasive, potentially risky, and more expensive than noninvasive g with ultrasound.

PCT/U52012/033584 Advanced imaging modes on ultrasound scanners have led to increased interest in imaging important quantities like wall shear rate (WSR) using Doppler (see, Blake, et al., “A method to estimate wall shear rate with a clinical ultrasound scanner,” Ultrasound in ne and y, vol. 34, no. 5, pp. 760-764, MAY 2008) and tissue/wall motion (see, Tsou et al., “Role ofultrasonic shear rate estimation errors in assessing inﬂammatory response and ar risk,” Ultrasound in Medicine and Biology, vol. 34, no. 6, pp. 963~972, JUN 2008; Karimi et al., “Estimation ofNonlinear . Mechanical Properties ofVascular Tissues via Elastography,” Cardiovascular Engineering, vol. 8, no. 4, pp. 191-202, DEC 2008; and Weitzel, et al., “High- tion Ultrasound Elasticity Imaging to Evaluate is Fistula Stenosis,” Seminars In Dialysis, vol. 22, no. 1, pp. 84-89, JAN—FEB 2009) using speckle tracking.

Recently, there has been increased interest in imaging ﬂow in conjunction with computational ﬂuid dynamic (CFD) modeling the tion of large artery hemodynamics (see, Steinman et al., “Flow imaging and computing: Large artery hemodynamics,” ANNALS OF BIOMEDICAL ENGINEERING, vol. 33, no. 12, pp. 1704—1709, DEC 2005; oa, et al., “A computational framework for ﬂuid—solid- growth modeling in cardiovascular tions,” Computer Methods in Applied Mechanics and Engineering, vol. 198, no. 45- 46, pp. 3583 — 3602, 2009; and Taylor et al., “Open problems in computational vascular biomechanics: Hemodynamics and arterial wall mechanics,” Computer Methods in Applied Mechanics and Engineering, vol. 198, no. 45—46, pp. 3514 — 3523, 2009). In this t, modeling ﬂuid-solid interfaces has been deﬁned as a challenge area in vascular mechanics.

SUMIVIARY At least one embodiment ofthis disclosure relates to ultrasound g capable of simultaneously imaging both wall tissue motion (e.g., perivascular tissue) and deformation, together with ﬂuid flow. For example, in one ment ofthis disclosure, imaging vascular mechanics using ultrasound is lished utilizing speckle tracking (e.g., a 2D phase coupled speckle tracking method suitable for subsample displacement estimation in both axial and lateral directions with minimum interpolation) in conjunction with an imaging mode (e.g., M2D mode g) that provides sufﬁcient frame rates for vector displacement tracking in both tissue and ﬂuid PCT/U52012/033584 simultaneously. For example, MZD imaging may be implemented on a clinical scanner equipped with a research interface for controlling the imaging sequence and streamlining the RF data for performing 2D speckle tracking in a region of interest (e.g., around a blood vessel). Combining 2D speckle tracking with sufﬁciently high frame rate imaging allows for ﬁne displacement tracking in both lateral and axial directions. Vector displacement ﬁelds resulting from such processing are well suited for strain and shear strain calculations with minimum ﬁltering and using relatively small ng s (i.e., speckle s) to maximize resolution. Flow and tissue motion strain ﬁelds (e.g., in a tissue/ﬂow application, such as in vivo in the carotid of a patient) may be evaluated (e.g., to identify vascular characteristics or for one or more other es, e.g., for use in therapy).

One exemplary embodiment of an imaging method may include providing ultrasound pulse—echo data of a region in which at least one portion of a blood vessel is located (e.g., wherein the pulse—echo data comprises pulse-echo data at a frame rate such that measured displacement of the vessel wall deﬁning the at least one portion ofthe blood vessel and measured average blood ﬂow through the at least one portion ofthe blood vessel have a quasi—periodic proﬁle over time to allow motion tracking of both the vessel wall and the blood ﬂow aneously) and generating strain and shear strain image data for the region in which the at least one portion of the vessel is located using e tracking. For example, the speckle tracking may e using multi— dimensional correlation ofpulse—echo data of one or more speckle regions undergoing deformation in the region in which the at least one portion of a blood vessel is located (e.g., wherein the multi—dimensional correlation ses determining a cross— correlation peak for the sampled pulse—echo data based on phase and magnitude nts of the cross-correlated pulse-echo data). Further, the method may e fying at least one vascular characteristic ofthe region in which at least one portion of a blood vessel is located based on the strain and shear strain image data (e.g., wherein the at least one vascular characteristic comprises at least one of a ﬂow characteristic associated with ﬂow through the blood , a structural characteristic ated with the blood vessel, and a hemodynamic characteristic associated with the blood vessel).

Another exemplary imaging method may include providing ultrasound pulse—echo data of a region in which at least one portion of a blood vessel is located and using PCT/U52012/033584 speckle tracking of one or more speckle regions ofthe region in which at least one portion ofthe blood vessel is located to track motion of both the vessel wall deﬁning the at least one portion ofthe blood vessel and the blood flow through the at least one portion ofthe blood vessel. The pulse—echo data is provided at a frame rate such that displacement of the vessel wall deﬁning the at least one n of the blood vessel and blood ﬂow through the at least one portion ofthe blood vessel are measurable simultaneously within a same periodic cycle (e.g., corresponding to a cardiac pulse cycle). r, the method may include identifying at least one vascular characteristic ofthe region in which at least one portion of a blood vessel is located based on the simultaneously ed displacement ofthe vessel wall and average blood ﬂow. Such an imaging method may also include generating strain and shear strain image data for the region in which the at least one portion ofthe vessel is located using the speckle tracking, wherein the speckle tracking comprises using multi—dimensional ation of sampled echo data of the one or more speckle regions undergoing deformation in the region in which the at least one portion of a blood vessel is located (e.g., the multi— dimensional correlation may include determining a cross-correlation peak for the sampled pulse-echo data based on phase and magnitude gradients ofthe cross—correlated d pulse-echo data).

In one or more embodiments of methods bed , identifying at least one vascular characteristic may include identifying one or more vessel wall ries; and still speckle tracking of such methods may include modifying a characteristic of at least one ofthe one or more e s being tracked (e.g., location, size, shape, etc.) based on the one or more vessel wall boundaries identiﬁed such that the at least one speckle region is entirely within or outside ofthe vessel wall.

Further, in one or more embodiments of methods described herein, identifying at least one vascular characteristic may include identifying vessel wall boundaries around the entire blood vessel (e.g., wall boundaries in the entire cross—section view taken along the axis of the vessel); may include measuring tissue property within the one or more vessel wall boundaries (e.g., ess or compliance); may include identifying one or more portions of a plaque architecture adjacent the one or more vessel wall boundaries (e.g., such that therapy may be focused on a portion of such a structure; such as the base thereof); and/or may include ating one or more hemodynamic measurements based PCT/U52012/033584 on both the motion tracking motion ofthe vessel wall and the blood ﬂow aneously.

Still timber, in one or more embodiments of methods described herein, using multi-dimensional correlation of d pulse-echo data of one or more speckle regions may e using two~dimensional correlation of sampled pulse—echo data of one or more speckle regions (tag, to track wall displacement or blood ﬂow), and even three- dimensional ation.

Yet further, in one or more embodiments of methods described herein, the method may ﬁarther include delivering therapy to a patient based on the identiﬁcation of the at least one vascular characteristic of the region in which at least one portion of a blood vessel is located (e.g., using ultrasonic energy to deliver y based on the identiﬁcation ofthe at least one vascular characteristic of the region in which at least one portion of a blood vessel is located). For example, at least one transducer conﬁgured to transmit and receive ultrasonic energy may be provided, wherein the at least one ucer is used to obtain the pulse—echo data (e.g., for image data generation) and to generate ultrasonic energy to deliver therapy.

In one or more other ments of methods described herein, generating strain and shear strain image data for the region in which the at least one portion of the vessel is located using two~dimensional speckle tracking may include generating at least one of axial strain and axial shear strain image data and/or lateral strain and lateral shear strain image data. Further, in such methods, ing ultrasound pulse-echo data of a region in which at least one portion of a blood vessel is located may include using coded excitation.

In yet one or more other embodiments of methods described herein, the method may include applying a dereverberation ﬁlter to the pulse—echo data from one or more speckle s in the blood to remove echo ents in the pulse—echo data due to reﬂection at the vessel wall when performing speckle tracking ofthe pulse-echo data from the one or more speckle regions in the blood.

PCT/U52012/033584 Another exemplary imaging method may include providing ultrasound pulse~echo data of a region in which at least one portion of a blood vessel is located; using e tracking of one or more speckle regions ofthe region in which the at least one portion of the blood vessel is d to track motion of at least one of the vessel wall defining the at least one portion of the blood vessel and the blood flow through the at least one portion ofthe blood vessel; identifying one or more vessel wall boundaries based on the speckle tracking of the one or more speckle regions; and modifying at least one characteristic of at least one of the one or more speckle regions being tracked based on the one or more vessel wall boundaries identiﬁed such that the at least one speckle region is entirely within or e ofthe vessel wall (e.g., the at least one of the one or more speckle s being tracked may be modiﬁed by at least one of location, size, or shape based on the one or more vessel wall boundaries ﬁed such that the at least one speckle region is entirely within or outside ofthe vessel wall).

In another exemplary imaging method, the method may include ing ultrasound pulse-echo data of a region in which at least one portion of a blood vessel is located; using speckle tracking of one or more speckle regions ofthe region in which at least one portion of the blood vessel is d to track motion of at least blood ﬂow through the at least one portion ofthe blood vessel; and removing echo ents in the pulse—echo data due to reﬂection at the vessel wall when performing speckle tracking ofthe pulse-echo data from the one or more speckle regions in the blood (e.g., ng echo components in the pulse—echo data due to reﬂection at the vessel wall may include using a time—varying inverse ﬁlter to reduce the components in the pulse—echo data due to reﬂection at the vessel wall).

One exemplary embodiment of a system for vascular imaging may include one or more ultrasound transducers (e.g., wherein the one or more transducers are red to deliver ultrasound energy to a vascular region resulting in pulse—echo data therefrom) and processing apparatus red (e.g., operative by execution of one programs, routines, or instructions to cause the performance of one or more functions) to control the capture of pulse—echo data at a frame rate such that measured displacement of a vessel wall deﬁning at least one portion of a blood vessel in the vascular region and measured average blood flow through the at least one portion ofthe blood vessel have a quasi—periodic proﬁle over time to allow motion tracking of both the vessel wall and the 2012/033584 blood flow simultaneously; to generate strain and shear strain image data for the region in which the at least one portion of the vessel is located using speckle tracking (e.g., wherein the speckle tracking may e using multi—dimensional correlation of pulse- echo data of one or more speckle regions undergoing deformation in the region in which the at least one portion of a blood vessel is located; the multi-dimensional correlation may include determining a cross—correlation peak for the d pulse—echo data based on phase and magnitude gradients ofthe cross—correlated pulse—echo data); and to identify at least one vascular teristic ofthe vascular region in which at least one portion of a blood vessel is located based on the strain and shear strain image data (e.g., wherein the at least one vascular teristic comprises at least one of a ﬂow characteristic associated with ﬂow through the blood vessel, a structural characteristic associated with the blood vessel, and a hemodynamic characteristic associated with the blood vessel).

Another exemplary system for vascular imaging may include one or more ultrasound transducers (e.g., wherein the one or more transducers are conﬁgured to deliver ultrasound energy to a vascular region resulting in pulse—echo data thereﬁom) and processing apparatus red (cg, operative by execution of one programs, routines, or instructions to cause the performance of one or more functions) to control the capture of pulse-echo data of the vascular region in which at least one portion of a blood vessel is located and use speckle tracking of one or more e regions of the vascular region in which at least one portion ofthe blood vessel is located to track motion ofboth the vessel wall deﬁning the at least one portion ofthe blood vessel and the blood flow through the at least one portion of the blood . The pulse-echo data may be captured at a frame rate such that displacement ofthe vessel wall deﬁning the at least one portion ofthe blood vessel and blood ﬂow through the at least one portion of the blood vessel are measurable simultaneously within a same periodic cycle corresponding to a cardiac pulse cycle. Further, the processing tus may be conﬁgured to identify at least one vascular characteristic ofthe vascular region in which the at least one portion ofthe blood vessel is located based on the simultaneously measured displacement of the vessel wall and average blood ﬂow. In one embodiment of such a system, the processing apparatus may further be operable to te strain and shear strain image data for the region in which the at least one portion of the vessel is located using the e tracking (e.g., wherein the e tracking may include using PCT/U52012/033584 multi—dimensional correlation of sampled pulse—echo data ofthe one or more speckle regions undergoing deformation in the region in which the at least one portion of a blood vessel is located; and further wherein the multi-dimensional correlation may include ining a cross—correlation peak for the d pulse—echo data based on phase and magnitude gradients ofthe cross-correlated sampled pulse-echo data).

In one or more ments ofthe exemplary systems provided herein, the processing apparatus may be operable to identify one or more vessel wall ries, and still further, the processing apparatus may be operable, when using the speckle tracking, to modify a characteristic of at least one of the one or more speckle regions being tracked (e.g., the location, size, shape, etc.) based on the one or more vessel wall boundaries identiﬁed such that the at least one speckle region is ly within or outside ofthe vessel wall.

Further, in one or more embodiments of the exemplary systems, the processing apparatus may be le to fy vessel wall boundaries around the entire blood vessel; the processing apparatus may be operable to e tissue property within the one or more vessel wall boundaries; the processing apparatus may be operable to identify one or more portions of a plaque architecture nt the one or more vessel wall ries; and/or the processing tus may be operable to calculate one or more hemodynamic measurements based on both the motion tracking motion ofthe vessel wall and the blood flow simultaneously.

Still further, in one or more embodiments of the exemplary systems provided , the processing apparatus may be operable to use two-dimensional correlation of sampled pulse-echo data of one or more speckle regions (e.g., to track speckle regions), and even three~dimensional correlation.

Yet timber, in one or more embodiments ofthe exemplary systems provided herein, the system may further include a therapy apparatus to deliver therapy to a patient based on the identiﬁcation of the at least one vascular characteristic ofthe region in which at least one portion ofa blood vessel is located (e.g., a device operable to use ultrasonic energy to deliver therapy based on the identiﬁcation ofthe at least one vascular characteristic of the region in which at least one portion of a blood vessel is PCT/U52012/033584 located). For example, the therapy tus may include at least one transducer conﬁgured to transmit and receive ultrasonic energy, wherein the at least one transducer is operable to e ultrasonic energy to deliver y based on the ﬁcation of the at least one vascular characteristic of the region in which at least one portion of a blood vessel is located and the at least one transducer is operable for use in obtaining the pulse~echo data to generate image data.

Still r, in one or more embodiments of the exemplary systems provided herein, the processing apparatus may be operable to te strain and shear strain image data for the region in which the at least one portion of the vessel is located using mensional speckle tracking, wherein using mensional speckle tracking comprises generating at least one of axial strain and axial shear strain image data and/or lateral strain and lateral shear strain image data. Further, for example, the processing apparatus may be operable to control providing ultrasound pulse-echo data of a region in which at least one n of a blood vessel is located using coded excitation.

Still r, in another ofthe one or more embodiments of the exemplary systems provided herein, the processing tus may be operable to apply a dereverberation ﬁlter to the pulse—echo data from one or more speckle regions in the blood to remove echo components in the pulse—echo data due to reﬂection at the vessel wall when performing speckle tracking of the pulse—echo data from the one or more speckle regions in the blood.

Another exemplary system for vascular imaging may include one or more ound transducers (eg, wherein the one or more transducers are conﬁgured to deliver ultrasound energy to a vascular region resulting in pulse—echo data therefrom) and processing apparatus conﬁgured to control capture of ultrasound pulse—echo data of the vascular region in which at least one portion of a blood vessel is located; use speckle tracking of one or more e regions ofthe vascular region in which the at least one portion ofthe blood vessel is located to track motion of at least one ofthe vessel wall deﬁning the at least one portion of the blood vessel and the blood ﬂow through the at least one portion of the blood vessel; identify one or more vessel wall boundaries based on the speckle tracking ofthe one or more speckle regions; and modify a characteristic of at least one of the one or more speckle regions being tracked based on the one or more PCT/U52012/033584 vessel wall boundaries identiﬁed such that the at least one speckle region is entirely within or outside ofthe vessel wall. For example, the processing apparatus may be operable to modify at least one of on, size, or shape of the at least one e region based on the one or more vessel wall boundaries identiﬁed such that the at least one speckle region is entirely within or outside ofthe vessel wall.

In yet another exemplary system for ar imaging, the system may include one or more ound transducers (e.g., wherein the one or more transducers are conﬁgured to deliver ultrasound energy to a vascular region resulting in pulse-echo data therefrom) and processing apparatus red to control e of ultrasound pulse—echo data of the vascular region in which at least one portion of a blood vessel is located; use speckle tracking of one or more speckle regions of the region in which at least one n of the blood vessel is located to track motion of at least blood flow through the at least one portion ofthe blood vessel; and remove echo components in the pulse-echo data due to reﬂection at the vessel wall when performing e tracking of the pulse-echo data ﬁom the one or more speckle regions in the blood (e.g., using a time—varying inverse ﬁlter to reduce the components in the pulse—echo data due to reﬂection at the vessel wall).

Still further, another exemplary system for vascular imaging may include one or more ultrasound transducers (e.g., wherein the one or more transducers are conﬁgured to deliver ultrasound energy to a vascular region resulting in pulse-echo data therefrom); apparatus for controlling the capture of pulse—echo data at a frame rate such that measured displacement of a vessel wall deﬁning at least one portion of a blood vessel in the vascular region and measured average blood ﬂow through the at least one n of the blood vessel have a quasi—periodic proﬁle over time to allow motion tracking of both the vessel wall and the blood ﬂow simultaneously; apparatus for generating strain and shear strain image data for the region in which the at least one portion ofthe vessel is located using speckle tracking (e.g., wherein the speckle tracking may include using inulti-dimensional ation of pulse—echo data of one or more e regions undergoing deformation in the region in which the at least one n of a blood vessel is located, and further wherein the multi—dimensional correlation may include determining a cross—correlation peak for the d pulse—echo data based on phase and magnitude gradients of the cross—correlated pulse—echo data); and apparatus for -10..

PCT/U52012/033584 identifying at least one vascular characteristic ofthe vascular region in which at least one portion of a blood vessel is located based on the strain and shear strain image data (e.g., wherein the at least one vascular characteristic comprises at least one of a ﬂow characteristic associated with ﬂow through the blood vessel, a structural characteristic associated with the blood vessel, and a hemodynamic characteristic associated with the blood vessel). Further, for example, the system may include therapy apparatus for delivering therapy to a patient based on the identiﬁcation ofthe at least one vascular characteristic of the region in which at least one portion of a blood vessel is d (e.g., ultrasound therapy tus).

AThe above summary is not ed to describe each embodiment or every implementation of the present disclosure. A more complete understanding will become apparent and appreciated by referring to the following detailed description and claims taken in conjunction with the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS is a block diagram depicting an exemplary ultrasound imaging , with an optional therapy .

I31] is a ﬂow chart depicting an exemplary ultrasound imaging method. is a block diagram of one exemplary embodiment ofan g system shown generally in Figure 1. is a block m of one ary GPU implementation of an imaging system such as shown in Figure 3.

FIG. SA-SD provides exemplary graphs showing channel diameter and average ﬂow velocity over time for image data captured at various frame rates.

FIGS. 6A—6B shown graphs including contours of certain parameters of cross— correlations used to describe one exemplary embodiment of speckle ng that may be used in an imaging method and/or system shown generally in Figures 1-2.

PCT/U52012/033584 provides an exemplary image of a blood vessel for use in describing one or more methods and/or systems shown generally in s 1—2 as they relate to vascular diagnostics or vascular therapy.

FIGS. 8A-SB show axial stain and axial shear strain images of ﬂow channel walls ng to examples carried out and bed at least in part herein.

FIGS. 9A—9B show l stain and lateral shear strain images ofﬂow channel walls relating to examples carried out and described at least in part herein. shows a graph of channel diameter computed from tracked channel wall displacements over time and average ﬂow velocity obtained from tracked ﬂuid displacements over time Within a ﬂow channel relating to an example carried out and described at least in part herein. shows graphs oftotal displacement vector waveforms with a channel at ent time ces relating to an example carried out and described at least in part herein.

FIGS. B show axial stain and axial shear strain images of carotid artery longitudinal vessel walls relating to examples carried out and described at least in part herein.

FIGS. l3A-13B show lateral stain and lateral shear strain images of carotid artery longitudinal vessel walls relating to examples carried out and described at least in part FIGS. 14A-14B Show axial stain and axial shear strain images of d artery cross-sectional vessel walls relating to examples carried out and described at least in part herein.

FIGS. ISA—15B show lateral stain and lateral shear strain images of carotid artery cross~sectional vessel walls relating to examples carried out and described at least in part herein.

PCT/U52012/033584 FIGS. 16A-l 6B show lateral and axial displacements relating to dereverberation ﬁltering examples. [46} FIGS. 17A—l7B show spatio—temporal maps relating to dereverberation ﬁltering examples.

I47] FIGS. B show spatio—temporal maps relating to dereverberation ﬁltering FIGS. 19A—19D show graphs relating to dereverberation ﬁltering results of examples presented herein.

DETAILED PTION OF EXEMPLARY EMBODIMENTS I49] In the following detailed description of illustrative embodiments, reference is made to the accompanying ﬁgures ofthe drawing which form a part hereof, and in which are shown, by way of illustration, speciﬁc embodiments which may be practiced. It is to be understood that other embodiments may be utilized and structural changes may be made without departing from (e.g., still falling within) the scope ofthe disclosure presented hereby.

Exemplary methods, apparatus, and systems shall be described with reference to Figures 1—19. It will be apparent to one skilled in the art that elements or processes (e.g., ing steps thereof) from one embodiment may be used in combination with ts or processes ofthe other embodiments, and that the le embodiments of such methods, apparatus, and systems using combinations of features set forth herein is not limited to the speciﬁc embodiments shown in the Figures and/or bed herein.

Further, it will be recognized that the embodiments described herein may include many elements that are not necessarily shown to scale. Still ﬁirther, it will be recognized that timing ofthe processes and the size and shape of various elements herein may be d but still fall within the scope of the present disclosure, although certain timings, one or more shapes and/or sizes, or types of elements, may be ageous over others.

PCT/U52012/033584 shows an exemplary ultrasound imaging system 10 including processing apparatus (block 12) and one or more ultrasound transducers, such as a transducer array that provides for transmission of pulses and reception of echoes (block 22). The sing apparatus (block 12) may be operably coupled to the one or more transducers (block 22) to facilitate imaging ofan object of interest (e.g., capture of pulse-echo data) using the one or more transducers (block 22). Further, the sing tus (block 12) includes data storage (block 14). Data storage (block 14) allows for access to sing programs or routines (block 16) and one or more other types of data (block 18) that may be employed to carry out the exemplary g methods (e.g., one which is shown generally in the block diagram of.

For example, processing programs or routines (block 16) may include programs or routines for performing computational mathematics, matrix mathematics, compression thms (e.g., data compression algorithms), calibration algorithms, image construction algorithms, inversion algorithms, signal processing algorithms, rdization algorithms, comparison thms, vector mathematics, or any other processing required to implement one or more embodiments as described herein (e.g., provide imaging, carry out speckle tracking, generate strain images, etc). ary mathematical formulations/equations that may be used in the systems and methods described herein are more speciﬁcally described herein with reference to FIGS. 3—19.

Data (block 18) may include, for example, sampled pulse—echo information (e.g., sampled or collected using the one or more transducers (block 22)), data representative ofmeasurements (e.g., vascular characteristics), results from one or more processing programs or routines employed according to the disclosure herein (e.g., reconstructed strain images of an object of interest, such as a blood vessel or regions around same), or any other data that may be ary for carrying out the one or more processes or methods bed herein.

In one or more embodiments, the system 10 may be implemented using one or more computer programs executed on programmable ers, such as computers that include, for example, processing capabilities (e.g., er processing units (CPUS), graphical processing units (GPUs)), data storage (e.g., volatile or latile memory and/or storage elements), input devices, and output devices. Program code and/or logic -14_ PCT/U52012/033584 described herein may be applied to input data to perform functionality described herein and generate desired output ation (e.g., strain images, vascular characteristics, etc.). The output information may be applied, or otherwise used, as input to, or by, one or more other devices and/or processes as described herein (e.g., one or more therapy tus (block 20) such as a drug therapy apparatus, an ultrasound therapy apparatus, etc.).

The program(s) or routine(s) used to ent the ses described herein may be provided using any programmable language, e.g., a high level procedural and/or object orientated programming language that is suitable for communicating with a computer system. Any such programs may, for example, be stored on any suitable device, e.g., a storage media, readable by a general or special purpose program, computer or a processor apparatus for conﬁguring and operating the er (e. g., processor(s)) when the suitable device is read for performing the procedures described herein. In other words, at least in one embodiment, the system 10 may be implemented using a computer readable storage medium, conﬁgured with a er m, where the storage medium so conﬁgured causes the computer to operate in a speciﬁc and predeﬁned manner to perform functions described herein.

Likewise, the imaging system 10 may be conﬁgured at a remote site (e.g., an application server) that allows access by one or more users via a remote computer apparatus (e.g., via a web browser), and allows a user to employ the functionality according to the present disclosure (e.g., user accesses a graphical user interface associated with one or more programs to process data).

The processing tus (block 12), may be, for example, any ﬁxed or mobile computer system (e.g., a personal computer or minicomputer, for example, with a CPU, GPU, etc.). The exact conﬁguration ofthe ing apparatus is not ng and essentially any device capable of providing le computing capabilities and l capabilities (e.g., control the imaging set up conﬁguration and acquire data, such as pulse—echo data) may be used. Further, various peripheral devices, such as a computer y, mouse, keyboard, , printer, scanner, etc. are contemplated to be used in combination with the processing apparatus (block 12), such as for visualization of 2012/033584 imaging results (e.g., display of strain images, display oftherapy delivery in real time such as with use of high intensity focused ultrasound, etc).

Further, in one or more embodiments, the output (e.g., an image, image data, an image data ﬁle, a digital ﬁle, a file in user—readable format, etc.) may be ed by a user, used by another machine that provides output based thereon, etc.

I59] As described herein, a digital ﬁle may be any medium (e.g., volatile or non—volatile memory, a CD—ROM, a punch card, magnetic recordable tape, etc.) containing digital bits (e.g., encoded in binary, trinary, etc.) that may be readable and/or writeable by sing tus (block 14) described herein.

Also, as bed herein, a file in user—readable format may be any representation of data (cg, ASCII text, binary numbers, hexadecimal numbers, decimal numbers, audio, graphical) presentable on any medium (e.g., paper, a display, sound waves, etc.) readable and/or understandable by a user. [61l lly, the methods and systems as described herein may utilize algorithms implementing computational mathematics (e.g., matrix. inversions, tutions, Fourier transform techniques, etc.) to reconstruct the images described herein (e.g., from pulse- echo data).

In View of the above, it will be readily apparent that the functionality as described in one or more embodiments according to the present disclosure may be implemented in any manner as would be known to one skilled in the art. As such, the computer language, the computer system, or any other re/hardware which is to be used to implement the processes described herein shall not be limiting on the scope of the systems, processes or ms (e.g., the functionality provided by such systems, processes or programs) described herein.

I63] One will recognize that a graphical user interface may be used in conjunction with the embodiments described herein. The user ace may provide various features allowing for user input thereto, change of input, importation or exportation of ﬁles, or any other features that may be lly suitable for use with the processes described —16— PCT/U52012/033584 herein. For example, the user interface may allow default values to be used or may require entry of certain values, , threshold , or other pertinent information.

The methods described in this disclosure, including those uted to the systems, or various constituent components, may be implemented, at least in part, in hardware, software, ﬁrmware, or any combination thereof. For example, various aspects ofthe ques may be implemented within one or more processors, including one or more microprocessors, DSPs, ASICS, FPGAs, or any other equivalent integrated or discrete logic circuitry, as well as any ations of such components, image processing s, or other devices. The term "processor" or "processing try" may generally refer to any ofthe foregoing logic circuitry, alone or in combination with other logic circuitry, or any other equivalent circuitry.

Such hardware, software, and/or e may be implemented within the same device or within separate devices to support the various operations and functions described in this disclosure. In addition, any of the described components may be implemented together or separately as te but interoperable logic devices.

Depiction of different features, e.g., using block diagrams, etc., is intended to highlight different functional s and does not necessarily imply that such features must be realized by separate hardware or software components. Rather, functionality may be performed by separate hardware or software ents, or integrated within common or separate hardware or software components.

When implemented in software, the functionality ascribed to the s, devices and methods described in this sure may be embodied as instructions on a computer—readable medium such as RAM, ROM, NVRAM, EEPROM, FLASH memory, magnetic data storage media, optical data storage media, or the like. The instructions may be executed by one or more processors to support one or more aspects ofthe ﬁinctionality described in this disclosure.

The imaging system 10 may further be used with, or may form a part of an optional therapy apparatus (block 20). For example, the therapy tus (block 20) may use the results of ultrasound imaging to provide one or more therapies. In one or more embodiments, the therapy apparatus (block 20) may be a non-invasive or invasive PCT/U52012/033584 therapy tus such as a drug delivery apparatus or system (delivery of a drug to a particular location), a surgical apparatus or system (e.g., delivery of a stent to a particular position), an ablation apparatus or system (e.g., a high frequency or high intensity d ultrasound therapy apparatus or ), etc.

I68] In one or more embodiments, the therapy apparatus (block 20) may be a separate system or apparatus that receives an output from the g system (e.g., image information) and delivers one or more ies. In other embodiments, the y apparatus (block 20) may be integrated with the imaging system to perform the one or more therapies (e.g., a high intensity focused ultrasound system that uses dual mode ultrasound transducer(s); for diagnostics such as imaging, as well as for treatment, such as ablation). For example, in one or more embodiments, the therapy apparatus (block 20) may include one or more portions of a system such as described in PCT International Publication No. W02009/002492 entitled “Image Guided Plaque Ablation,” published 31 er 2008, and incorporated herein by reference. For example, the ound imaging described herein may be used for ng vascular plaque non—invasively. For example, the ultrasound imaging described herein may be used to identify flow and vascular characteristics needed to non—invasively perform ablation of plaque as described in PCT International Publication No. W02009/002492.

For example, the therapy system may be a system for non—invasively elevating the temperature oftissue by ultrasound energy waves including: at least one ultrasound delivery device adapted to deliver ultrasound energy waves to a focal point oftargeted tissue; a ature monitoring device for monitoring the temperature oftargeted tissue at the focal point; and a controller for steering and controlling the ultrasound delivery device to deliver ound energy waves at a focal point to elevate the ature of targeted tissue to a desired ature. [701 Further, for example, the therapy system may use one or more imaging systems described herein to produce an image of at least a portion of a mammalian body, e.g., such that the location of at least one vascular plaque in said image can be determined and to ascertain the location ofthe base of said vascular plaque. For example, ultrasound delivery device may ain one or more target locations at the base ofthe plaque. Still further one or more embodiments of the imaging system provided herein may be used in ~18— PCT/U52012/033584 a method for elevating the temperature at a target location by an energy wave using an ultrasound therapy system (e.g., which may be the same ultrasound system (ultrasound transducers thereof) used for imaging). For example, the method may include delivering a beam of ultrasound energy waves from a source to the target location; monitoring the temperature ofthe target location; and stopping the delivering of the beam ofultrasound energy waves if a desired temperature at the target location has been reached.

Further, a method of preparing a plan for non—invasively elevating the ature oftissue in a vessel wall leading to regression of ar plaques may include imaging at least a portion of a body to produce an image (e.g., using ultrasound imaging as described herein to image a vascular region); determining the location of at least one ar plaque in said image; ascertaining the location of the base of said vascular plaque and one or more target ons at the base ofthe plaque (e.g., using the ound generated image); and/or determining the parameters for delivering ound energy waves from a source to a focal point for elevating the ature of targeted tissue in the vessel wall to a desired temperature, sufﬁcient for reducing or destroying vaso vasorum.

Further, for example, the ultrasound imaging described herein may be used to fy ﬂow and vascular characteristics needed to perform invasive treatments of plaque (e.g., stent ry, c surgery, etc.) Still further, in one or more embodiments, the therapy apparatus (block 20) may include one or more portions of a system such as described in US Provisional Patent Application No. 61/353,096, entitled “Dual Mode Ultrasound Transducer (DMUT) System for Monitoring and Control of Lesion Formation Dynamics” ﬁled 9 June 2010, and which is incorporated by reference herein. For example, the ultrasound g described herein may be performed with the same or similar transducer arrays described therein which can be used for both imaging (e.g., to monitor a therapy procedure), as well as for delivering y (e.g., to deliver high intensity focused ultrasound energy).

For example, therapy may be red using the ultrasound ucer array, while the imaging modes using the same transducer array may be used to guide the therapeutic beam, assess thermal and mechanical tissue response to estimate doses of therapy (e.g., initial dose oftherapy), monitor and characterize tissue response during therapy, and PCT/U52012/033584 assess the state ofthe treated tissue at the completion of each re to the therapeutic ound energy (e.g., real time monitoring n periods of therapy delivery).

For e, ultrasound imaging as described herein may be used to identify one or more vascular teristics. An exemplary diagram of a blood vessel 50 is shown in Figure 7 to facilitate discussion of the use of imaging described herein. The blood vessel 50 shown in Figure 7 includes a vessel wall 52 having a plaque structure 54 formed on the interior of the vessel wall 52. The plaque architecture ofthe structure 54 may include, for example, a plaque base 56, a lipid core 58, and a ﬁbrous or calciﬁed cap 60.

Blood 62 ﬂows through the blood vessel 50 deﬁned by the vessel wall 52.

One or more embodiments ofmethods and/or systems described herein may be used to identify one or more vascular characteristics, e.g., ﬂow characteristics associated with the ﬂow through the blood vessel 50, structural characteristics associated with the blood vessel 50, and/or hemodynamic characteristics. For example, ﬂow characteristics may include ﬂow velocity, volume flow, wall shear stress, wall shear rate, etc.

For e, structural characteristics may include determining boundaries of the vessel wall (e.g., outer and inner boundaries, such as in a coordinate system), thickness ofthe vessel wall, measurement oftissue properties within the vessel wall (e.g., stiffness oftissue, such as, for example, it relates to a diseased state), differentiation of plaque from vessel wall, differentiation ofthe s components of plaque (e.g., differentiation ofbase from lipid core, differentiation ofbase from ﬁbrous cap, differentiation of lipid core from ﬁbrous cap, etc), etc. For example, in one or more embodiments, upon differentiation ofthe base from the ﬁbrous cap of the plaque architecture, treatment may be provided to ablate the base to reduce ﬁlrther plaque p or growth or provide treatment according to PCT International Publication No. /002492.

Still further, for example, hemodynamic characteristics may include calculated namic measurements, such as, for example, arterial pressure, cardiac output, arterial compliance, pulse wave velocity, etc. At least in one embodiment, such namic measurements may be determined based on parameters ng to both tracking ofthe blood flow and tracking of vessel wall motion or displacement. As such, PCT/U52012/033584 to obtain an accurate hemodynamic determination, the parameters or measurements relating to both tracking of the blood ﬂow and tracking of vessel wall motion or displacement must be determined simultaneously, or within a periodic cycle in which both can be determined (e.g., determined effectively). For example, compliance of the vessel may be based on both volume ﬂow which relates to tracking of blood ﬂow and re within the vessel which can be determined by tracking vessel displacement.

For example, accurate tion ofthe vessel diameter and estimation of the l ﬂow within the lumen with high frame rate imaging will allow for useﬁil measurement of the pulse wave velocity (PWV) noninvasively. For example, the time waveforms shown in, for e, Figures SC—SD can be plotted in phase space (volume ﬂow, QA, vs. vessel area, A). Volume ﬂow can be calculated from the ﬂow data, while the area can be obtained from the vessel wall movement. This measurement must be made during the ion—free part ofthe heart cycle in the form of a slope measurement of the form PWV=dQ/dA. With the adequately sampled time waveforms (e.g., using MZD mode imaging), the task of estimating the vessel wall motion and the ﬂow within the vessel can be accomplished.

In other words, both lateral ﬂow velocity and wall motion can be ted aneously thus providing pressure (through vessel diameter) and flow (through vector velocity)). Such measurements can provide the basis for namic computations that may be used in the assessment of vessel wall compliance, an important indicator of the health ofthe vessel as described herein. Further, as described herein, axial and lateral displacement ﬁelds are ehaved and allow for strain and shear strain calculations in both tissue and blood. Together with anatomical image information, these velocity/strain ﬁelds may e input for computational ﬂuid dynamic models, which may allow for inverse calculations suitable for the assessment of the health of the vasculature and nding tissue (e.g. detection and staging of atherosclerosis).

In one or more embodiments, the ultrasound—enabled quantitative imaging system may be used for assessment ofthe disease state in atherosclerotic blood vessels. For example, the imaging may be used for the direct estimation of the strain ﬁelds in the vicinity of the vessel walls. Such methods may mitigate the deleterious effects of local PCT/U52012/033584 deformations that could result in loss ofcorrelation, and which may render the ation—based speckle tracking approach useless in the Vicinity ofthe vessel wall.

Such deformations, depending on severity, could result in erroneous estimate in the velocity (and therefore strain) estimation or may even result in loss of accuracy.

A three—pronged approach to the problem of restoring the true velocity/strain estimates may include: 1) A ep algorithm for direct estimation ofthe velocity/strain components using a deformed model of the 2D RF data in the vicinity of the wall, 2) A tructive approach employing a forward computational ﬂuid dynamics (CFD) model as a regularization ﬁlter, and 3) A quantitative inverse reconstruction ofthe tissue mechanical properties using ultrasound—based ty/strain ﬁelds as observations. As described herein, simultaneous imaging of tissue motion and ﬂow with subsample accuracy in both axial and lateral directions may be implemented.

For e, such imaging may include using a phase—coupled 2D e tracking approach, which employs the true 2D x cross correlation to ﬁnd sub—pixel displacements in both axial and lateral directions. Further, a modiﬁed imaging sequence on a Sonix RP scanner to allow high ﬁame rate 2D data collection in a limited ﬁeld of View ng the region of interest (MZD-mode) may be used. Together with the robust 2D speckle tracking method, M2D imaging allows for capturing the full dynamics ofthe ﬂow and wall/tissue motion, even when the flow is primarily in the lateral direction (with t to the imaging beam). The ﬁne vector cement estimates in both axial and lateral directions are shown to allow for smooth and contiguous strain and shear strain calculations with minimal ﬁltering. The simultaneous imaging of the vector flow ﬁeld and the wall/tissue motion and the corresponding strains at high spatial and temporal sampling may provide a tool in ng the ﬂuid—solid interactions between the blood and blood vessel. Such an image-based modeling of the vessel response may allow for the prediction of the disease state and possible ion of the disease state.

Furthermore, the ation n the observation model and inverse reconstruction ofthe tissue properties in the vicinity ofthe vessel wall may allow for quantitative assessment of the plaque composition (e.g. lipid content or calciﬁcation).

This may provide a reliable noninvasive model for selecting treatment options based on probability ofrupture and other risk factors.

PCT/U52012/033584 In other words, the imaging bed herein may be used in conjunction with computational ﬂuid dynamic (CFD) modeling the evaluation of large artery hemodynamics. CFD has been shown to produce useful prediction oftime—varying, 3D ﬂow ﬁelds in large arteries with complex geometries. In this context, modeling ﬂuid- solid interfaces has been deﬁned as a nge area in vascular mechanics. Imaging methods as described herein capable of capturing both perivascular (and wall) tissue motion and deformations, together with ﬂuid flow may be used to s this issue.

Advances in MRI and other imaging modalities have led to increased interest in image— based, patient—specific CFD modeling to monitor disease progression. MRI has excellent soft tissue contrast that may allow the accurate capture of the tissue (solid) model. Although this may be an advantage over diagnostic ound, which does not offer the same level of deﬁnition for tissue boundaries and discrimination between tissue types, this tion may be mitigated, however, by the recent improvements in 3D image acquisition, both freehand and zed. Therefore, diagnostic ultrasound scanners may provide an attractive alternative for image-based CFD modeling. Based on 3D ultrasound and improved 2D velocity/strain imaging using MZD mode, processes for providing quantitative tissue property images in the vicinity of the vessel wall for the characterization ofthe e state may be ented. The imaging s described herein address the limitations of existing correlation—based s for velocity/strain estimation to restore the lost or artifact—ridden tes in the vicinity of the wall. Further, the integration of our velocity/strain estimation as an ation model in a dynamic, forward/inverse CFD—based model for the reconstruction ofthe ﬁeld/tissue property values consistent with the -Stokes equations may be accomplished. For example, the following may be developed: a two—step algorithm for direct strain estimation at the vessel wall using MZD—mode data; a regularized approach for the reconstruction of displacement/strain maps utilizing a d CFD model obtained from 3D Ultrasound (e.g., the forward model may provide a reconstruction ﬁlter to regularize the velocity/strain estimation obtained using the speckle tracking algorithm); and/or an inverse method for the reconstruction ofthe mechanical properties in the Vicinity of the vessel wall based on strain maps obtained using de data.

The one or more ultrasound transducers (block 22) may be any apparatus (e.g., transmitting, receiving components, etc.) capable of delivering ultrasound pulses and PCT/U52012/033584 sampling/collecting ultrasound echo energy contemplated to be used in ound imaging systems and in combination with processing apparatus (block 12) of the system . As used herein, such transducers may include a transmitting portion, e.g., to deliver pulse energy, and a receiving portion, e.g., to sample/collect echo or reﬂected energy, which may or may not be the same n. During the ultrasound imaging of a target (e.g., a blood vessel, such as a carotid artery, coronary artery, etc.), the one or more transducers (block 22) may be positioned relative to the target so as to be capable of delivering energy to the target resulting in reflected energy (also known as the resultant pulse—echo or echo energy) and also sampling the echo .

The one or more ucers (block 22) may include multiple transducers position separately from one another or may be a transducer array. In one or more embodiments, various arrays may have one or more beneﬁts over others. For example, in one or more embodiments, the transducer array may be a segmented concave transducer with multiple sub—apertures to insonify the vessel from multiple angles. This will allow for better deﬁnition of the vessel boundaries from more directions. At least one sub—aperture may be used in linear array or phased array mode for initial B-mode and strain imaging of the vessel. The driver of the ucer may be designed to drive the multiple sub—apertures with independent codes. Each sub-aperture may be a one-dimensional or two— dimensional array. Coded excitation may help improve both the data rates (e.g., provide higher frame rates) and echo quality (e.g., by reducing reverberations within the lumen).

The receiver may be a multichannel receiver with beamforming and/or pulse compression for coded excitation.

For example, various arrays and operation thereof, are bed in Ebbini, et al., Mode Ultrasound Phased Arrays for Image—Guided Surgery,” Ultrasound Imaging, vol. 28, pp. 65—82 ; Ballard, et al., “Adaptive Transthoracic Refocusing ofDual—Mode Ultrasound Arrays,” IEEE Transactions on Biomedical Engineering, vol. 57, no. 1, pp. 93—402 (Jan. 2010); and Wan et al., “Imaging with Concave Large— Aperture Therapeutic Ultrasound Arrays Using Conventional Synthetic—Aperture Beamforming,” IEEE Transactions on Ultrasound, Ferroelectrics, and ncy Control, vol. 55, no. 8, pp. l705~l7l8 t 2008), which are all hereby incorporated by reference herein.

PCT/U52012/033584 A ﬂow chart of an exemplary ultrasound imaging method 30 for vascular imaging is depicted in One will recognize that one or more of the blocks of functionality described herein may be d out using one or more programs or routines, and/or any other components of an imaging system (eg, the imaging system 10 of and/or therapy system (e.g., the therapy system 20 of . 1831 Generally, the method 30 provides for the capture of echo data at a sampled ﬂame rate (block 32). In one embodiment, ultrasound pulse—echo data is provided of a region in which at least one portion of a blood vessel is located. For example, the pulse— echo data may be pulse—echo data sampled at a frame rate such that measured displacement of the vessel wall deﬁning the at least one portion ofthe blood vessel and measured average blood ﬂow through the at least one portion ofthe blood vessel have a quasi-periodic proﬁle over time to allow motion tracking of both the vessel wall and the blood ﬂow simultaneously. r, the method es applying speckle tracking (block 34) to the pulse~echo data to allow, for example, the generation of strain and shear strain image data.

As set forth herein with respect to the system shown in Figure 1, one or more vascular characteristics, e.g., ﬂow characteristics associated with the ﬂow through the blood vessel 50, structural characteristics associated with the blood vessel 50, and/or hemodynamic characteristics, may be identiﬁed (block 36) based on the tracking of motion in the vessel wall and ﬂow. In one embodiment, due the simultaneous e of displacement ﬁelds in both the ﬂow and vessel wall, e.g., during a ic cycle, such as a cardiac cycle, one or more vascular characteristics which depend on ements resulting from or ng to both such types of displacements (e.g., such as hemodynamics) may be determined.

Still further, as shown in Figure 2, optionally the method 30 may include delivering therapy based on one or more vascular characteristics (block 38). For example, as described with respect to the system of Figure 1, delivery of y may take one or more different forms (e.g., drug, ablation, surgical, or any other invasive or non—invasive treatment).

PCT/U52012/033584 In one or more embodiments, the method may include MZD mode imaging designed to ze the lateral extent of the imaged region at sufﬁciently high frame rates to e the full dynamics ofthe vessel wall and the ﬂow within the vessel. MZD mode produces 2D beamformed RF echo data ﬁom a selected region ofthe ﬁeld of view (FoV) of a given probe. The region may be contiguous or comprised ofmore than one disjoint subsegments. As an example, on the SonixRP scanner (Ultrasonix, BC, Canada), an ary set of A-lines within the FoV can be used to form the MZD mode image with frame mode approximately MB/MMgp higher than B-mode imaging, where M3 and MW]; indicate the number of A—lines used to form B—mode and MZD—mode images, respectively.

As shown in Figure 4, M2D mode imaging may be enabled by creating a powerﬁil ned execution/ﬂow program architecture capable of employing a variety of computational resources for real-time implementation. In addition, the program architecture may allow the user to invoke additional computational resources available on the computer (or generally on the Internet) to achieve other computational tasks. The results from these computations can be integrated seamlessly with the program. For example, the beamformed RF echo data may be transferred in real time h a Gigabit ace to allow real—time 2D axial strain computations using GPU (or FPGA) using 1D speckle tracking. However, the beamformed RF data is available for additional processing using, for example, a pre—installed MATLAB engine. The MATLAB results can be imported back seamlessly to the MZD mode g program with minimum latency (e.g., after the completion of the MATLAB calculations). This capability may allow us to perform real—time 2D speckle tracking to enable strain and shear strain in the ty of the vessel wall, e.g., heavy—duty MATLAB-based ations are performed on a small Rol allowing for their incorporation in real time. In at least one embodiment, true 2D speckle tracking approaches may be implemented in real time as is tly the case with 1D speckle tracking. In this way, a pipelined program execution architecture may be implemented to support MZD imaging which allows us to reap the beneﬁts of powerful computational tools for the analysis ofthe vessel walls in quasi ime.

The high frame rate MZD mode preserves the correlation to produce well—behaved 2D displacement/velocity proﬁles to allow for robust strain ation. High quality 2D ) strain and shear strain ﬁelds produce views ofthe vessel wall boundaries on —26- PCT/U52012/033584 both sides of the vessel in the axial view. Further, they may also produce better deﬁnition of the wall in the lateral direction in the cross—sectional View. This may allow for measurements ofwall thickening, an early sign of atherosclerosis. 194] Still further, the high quality 2D ; i.e., over time) strain and shear strain will allow for tissue property measurements within the vessel wall, e.g. ess. Such tissue property measurements will allow for the characterization of the disease state and, given the high resolution, the plaque architecture (e.g., base, lipid core, and ﬁbrous or calciﬁed cap). Therapy or treatments may be red based on such information or such ation may be used during the delivery of such y (e.g., high ity focused ultrasound treatments that target the base of the plaque without damage to the cap or even the lipid core, ual determination ofthe response oftissue to therapy, such as between doses ofhigh intensity focused ultrasound, etc.).

The pulse~echo data may be provided using any imaging system (e.g., the imaging system 10 of, although one or more imaging systems may be advantageous over others. In one or more embodiments, the data acquisition may be performed by an imaging system 100 such as shown in Figures 3 and 4. For example, as shown in Figure 3, the g system 100 may be used to acquire and perform real—time processing of such acquired data. The imaging system 100 may include an ultrasound scanner 102 (e.g., a Sonix RP (Ultrasonix, Canada)) loaded with a program used for high frame rate pulse—echo data collection. The ultrasound scanner may include es such as chirp generation, waveform generation, data collection, and data transfer capabilities. [961 As shown in the exemplary embodiment, collected data may be streamlined to a controller PC 104 through, for example, Gigabit Ethernet for real—time data processing.

The data processing computer 104 can handle the intensive computations required by high—resolution (both l and temporal) speckle tracking and separable 2-D post— ﬁltering by utilizing a many—core graphics processing unit 106 (6g, GPU; nVlDlA, Santa Clara, CA). The ultrasound scanner 102 may operate in B mode for image guidance (e.g., producing a B-mode image) and in M2D mode for high-frame—rate data collection. The MZD mode achieves ame—rate imaging by limiting the number of scan lines to the region of st (ROI), as deﬁned by the user. MZD data may be used for e tracking. For example, in one embodiment, MZD-mcde acquisition with 10 PCT/U52012/033584 A lines per frame at 1000 fps may be performed. For example, a linear array probe (e.g., LAl4—5/3 8) may be used for data collection at a frame rate of 1000 fps in MZD mode by ng the number of scan lines to 10 and the imaging range to 40 mm. In one or more embodiments with coded excitation, M2D~mode may collect data at ﬁ‘ame rates from 2000 to 5000 frames per second or more.

In the embodiment shown in Figure 3, high ity focused ultrasound (HIFU) is also possible (e.g., for generating ound for treatment or subtherapeutic mechanical and/or thermal s). As such, a VirtexZPro (Xilinx, CA) ﬁeld-programmable gate array (FPGA) board 108 is dedicated for HIFU source and synchronized frame trigger tion. This implementation allows an interference-free data collection by brieﬂy silencing the HIFU generator while pulse-echo imaging is active.

Speckle tracking used for imaging herein relies on accurate estimation of incremental frame—to—frame time shifts, which are typically much smaller than RF—echo sampling period. In one embodiment, a 2-D complex correlation oftwo subsequent ﬂames of pulse-echo data is carried out. The real—time data processing engine 110 is based on a GPU 106 with a large number of cores (e.g., a GTX285 GPU with 240 processing cores and designed to take full advantage of its highly parallel architecture).

Implementation of real—time processing is supplemented with ated Performance Primitives (Intel) and Matlab tools 112, 114. Further, a research interface system 120 is provided for operator control.

Figure 4 shows an ary block diagram of a GPU-based implementation 130 for an imaging , such as shown in Figure 3. In one embodiment, for each processing stage, ﬁne—grained partition is performed for the algorithm in a data- independent manner so that all 240 sors are working efﬁciently on individual blocks of data.

For example, Figure 4 shows one exemplary embodiment of a real—time signal sing chain of rmed ultrasound data for strain imaging (e.g., shown for 1D, but which can be generalized to 2D or 3D). The real-time speckle tracking is performed in the axial direction, but axial strain and axial shear strain can be performed in real—time as a first step to identify the vessel boundaries (e.g., this has been tested in vivo on ~28— PCT/U52012/033584 vessels of various diameters (~1 mm in rat) and (~4 mm in swine)). The 2D strain calculations can be performed in a region—of—interest (ROI) around the identiﬁed blood vessel. Axial and lateral strains and corresponding shear strains can be calculated in real—time following this step. Furthermore, direct estimation of s and shear strains in the ate vicinity of the vessel wall may be provided. For both 1D and 2D versions ofthe strain calculations, additional sing in the time direction will ﬁirther deﬁne the vessel boundaries and the internal boundaries of the wall. For example, in the context of imaging atherosclerosis, this may yield the internal composition of the plaque.

As shown in Figure 4, on the CPU side, data can be come from Network Stack (e.g., experimental mode, where data is streamlined ﬁom SonixRP scanner) or Data File (e.g., review mode). The sed result can be ized with a designed U1 system (OpenGL based) or exposed to other commercial soﬁware for ﬁirther analysis (e.g., Matlab). The result can be also used in feedback control for ime ature control.

On the GPU side, Figures 4 shows a GPU-based entation ofthe algorithm described in (see, e.g., Simon, et al., imensional Temperature Estimation Using stic Ultrasound,” IEEE Transactions on Ultrasonics, Fen'oelecn'onics, and Frequency Control, vol. 45, no. 4, pp. 1088-1099, JULY 1998). The various blocks shown in Figure 4 have at least the following functionality: Hilbert transform: computes the analytic signal of the RF echo using an FIR Hilbert Transformer; Cross correlation/Phase projection/Accumulate: implements 1D version of speckle tracking; 2D Separable Filter: allows temperature estimation (e.g., thermal strain computation); Bilinear Interpolation: provides hardware accelerated olation for data Visualization; and Local Storage: provides data management in GPU domain.

In other words, in one or more embodiments, a system for vascular imaging is provided herein that includes one or more ultrasound transducers (e.g., wherein the one or more ucers are configured to r ultrasound energy to a vascular region resulting in pulse-echo data therefrom) and processing apparatus (e.g., including one or more programs able by one or more processors ofthe system to perform one or more functions thereof and as described herein, such as control of data frame acquisition, speckle tracking, visualization, image generation, characteristic identiﬁcation, etc) PCT/U52012/033584 In other words, the processing apparatus (e.g., GPU, CPU, etc.) may be conﬁgured (e.g., e under control of one or more programs) to, for example, control the capture of pulse-echo data at a ﬂame rate such that measured displacement of a vessel wall g at least one portion of a blood vessel in the vascular region and measured average blood ﬂow through the at least one portion of the blood vessel have a quasi— periodic profile over time to allow motion ng of both the vessel wall and the blood ﬂow simultaneously; generate strain and shear strain image data for the region in which the at least one n of the vessel is located using speckle tracking; and identify at least one vascular teristic of the vascular region in which at least one portion of a blood vessel is located based on the strain and shear strain image data (e.g., wherein the at least one vascular characteristic comprises at least one of a ﬂow teristic associated with ﬂow through the blood vessel, a structural characteristic associated with the blood vessel, and a hemodynamic characteristic associated with the blood vessel).

Further, for example, processing tus may be conﬁgured to use e tracking of one or more speckle s of the vascular region in which at least one portion ofthe blood vessel is located to track motion of both the vessel wall deﬁning the at least one portion of the blood vessel and the blood ﬂow through the at least one portion ofthe blood vessel (e.g., n the pulse—echo data is captured at a frame rate such that displacement of the vessel wall deﬁning the at least one portion of the blood vessel and blood ﬂow through the at least one portion ofthe blood vessel are measurable simultaneously within a same periodic cycle corresponding to a cardiac pulse cycle and/or identify at least one vascular characteristic of the vascular region in which the at least one portion of the blood vessel is located based on the simultaneously measured displacement of the vessel wall and average blood ﬂow. Further, for example, the processing apparatus is further le (e.g., by ing one or more programs) to generate strain and shear strain image data for the region in which the at least one portion ofthe vessel is located using the speckle tracking (e.g., wherein the speckle tracking includes using multi—dimensional correlation of sampled pulse—echo data ofthe one or more speckle regions undergoing deformation in the region in which the at least one portion of a blood vessel is located, wherein the multi—dimensional correlation comprises determining a cross—correlation peak for the sampled pulse—echo data based on phase and magnitude gradients ofthe cross-correlated sampled pulse-echo data).

PCT/U52012/033584 As shown in Figures SA—SD, in one or more embodiments, the ﬂame rate at which pulse echo data is acquired should be at least greater than 100 fps, and even greater than 200 fps. In at least one embodiment, the ﬂame rate is greater than 300 fps. The frame rate should be sufﬁciently high for reliable motion tracking in both tissue and blood onously (e.g., such that measurements relating to the same are relevant to the same time frame or periodic cycle). For example, in one embodiment, the MZD mode allows for providing a high ﬂame rate while maintaining the correlation at high levels to produce smooth and uous displacement /velocity ﬁelds (e.g., displacement of tissue, motion of blood ﬂow) to allow for robust strain and shear strain determinations.

The ﬂame rate for acquiring pulse—echo data should be adequate to provide displacement ﬁelds in both flow and tissue that are well~behaved. In one embodiment, such well—behaved cement can be identiﬁed by the ements of channel er related to tracking oftissue displacement (e.g., vessel wall displacement or motion) and average ﬂow velocity related to the tracking ofﬂow through the vessel (e.g., tracking ofblood ﬂow through the blood vessel) over time. As such, Figures SA—SD show graphs ofboth the tracking ofﬂow in a vessel (cg, represented by the change of average ﬂow velocity shown in solid lines) as well as the tracking of vessel tissue cement (e.g., represented by change in channel diameter over time shown by dashed lines) for multiple ﬂame rates. Figure 5A shows determinations for a ﬂame rate of40 fps; Figure SB shows determinations for a ﬂame rate of 81 fps; Figure 5C shows determinations for a ﬂame rate of 162 fps; and Figure 5D shows determinations for a ﬂame rate of 325 fps.

It is clear ﬂom Figures 5A and 513, that such low ﬂame rates (e.g., less than 100 fps) do not e measured displacement ofthe vessel wall deﬁning a blood vessel and measured average blood ﬂow h the blood vessel which are well behaved. In other words, measurements of flow and tissue displacement (e.g., vessel wall displacement) could not be accurately ed simultaneously. For example, as clearly shown in Figures 5A and SB, the blood ﬂow is shown to be rather random throughout a periodic cycle of the displacement of the vessel wall represented by the channel diameter information (e.g., ﬂom peak to peak of the vessel wall displacement). In other words, such blood ﬂow data is ridden with artifacts making an accurate flow determination PCT/U52012/033584 difﬁcult during the cycle. Even the vessel wall displacement information appears to include some cts.

However, discernible from s 5C and 5D, at higher frame rates (e.g., r than 100 fps) the measured displacement ofthe vessel wall deﬁning a, blood vessel and measured average blood ﬂow through the blood vessel are well d such that ements of flow and tissue displacement (e.g., vessel wall displacement) could be accurately measured synchronously (e.g., corresponding to the same time). For example, as clearly shown in Figures 5C and 5D, the blood ﬂow is shown to be much less random throughout a periodic cycle ofthe displacement of the vessel wall represented by the. channel diameter information (cg, ﬁom peak to peak ofthe vessel wall displacement).

In other words, the measured displacement of the vessel wall deﬁning a blood vessel and measured average blood ﬂow through the blood vessel have a quasi-periodic proﬁle over time which allows motion tracking of both the vessel wall and the blood ﬂow simultaneously. Note the strong peaks for both the channel diameter and average ﬂow velocity within the same periodic cycle (e.g., corresponding to the cardiac cycle).

As used herein, the term quasi—periodic proﬁle is meant to reﬂect a proﬁle that is substantially consistent over periodic cycles in the form of a regularized pattern, even though there will be some variation on a frame to frame basis, e.g., periodic cycles corresponding to cardiac cycles. For example, such a quasi—periodic proﬁle for flow ty may e strong peaks during each cycle indicating maximum ﬂow followed by ﬂow measurements that indicate little or no flow during the remainder ofthe cycle.

Further, for example, such a quasi—periodic proﬁle for channel diameter may include strong peaks during each cycle indicating maximum displacement of the vessel wall followed by measurements that indicate a relaxation ofthe vessel to a normal state during the remainder of the cycle. Such a ﬂame rate that results in a periodic proﬁle allows for tracking of ﬂow and vessel displacement simultaneously, or in other words, synchronously with each other (e.g., in phase with each other).

The frame rate may vary depending on various s. For example, the frame rate may be based on the vessel structure (e.g., carotid artery versus peripheral vein), timing ofthe periodic flow through the vessel (e.g., pulse cycle length), motion of the vessel structure (e.g., time for vessel to relax to normal); depth ofthe target vessel (e.g., deeper PCT/U52012/033584 vessels may be imaged at lower frame rates), use of coded excitation (e.g., coded excitation may allow for increased ﬂame rates), and the f—number of imaging focus (e.g., higher f-numbers may result in reduced lateral resolution), etc.

Still further, in one or more embodiments, higher ﬂame rates may be accomplished when coded excitation ultrasound is used. Such coded excitation ultrasound is bed, for example, in the literature and will not be discussed in detail herein. For example, one or more rative examples of coded excitation ultrasound which may be used in ation with the g method and/0r systems described herein are provided in Shen et al., “A New Coded-Excitation Ultrasound Imaging System———Part I: Basic Principles,” IEEE ctions on onics, Ferroelectrics, and Frequency Control, vol. 43, no. 1, pp. 131 —140, Jan. 1996); Shen et al., “A New Coded-Excitation Ultrasound Imaging System--—Part II: or Design,” IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, vol. 43, no. 1, pp. 141 ~148, Jan. 1996); and Shen et al., “Filter~Based Coded—Excitation System for High-Speed Ultrasound hnaging,” IEEE Transactions on Medical Imaging, vol. 17, no. 6, pp. 923— 934, Dec. 1998), which are all incorporated herein by reference.

In one or more embodiments, the frame rate may be greater than 100 fps, greater than 200 ips, greater than 300 fps, greater than 500 fps, greater than 1000 fps, and even greater than 5000 fps. In other embodiments, the ﬂame rate may be less than 5000 ips, less than 4000 fps, less than 3000 fps, less than 2000 fps, less than 1000 fps, less than 600 fps, less than 500 fps, less than 400 fps, less than 300 fps, or less than 200 fps. In at least one embodiment, the frame rate is within the range of 100 fps to 5000 fps.

As shown in the imaging method 30 in Figure 2, speckle tracking is applied to the pulse-echo data (block 34). For example, strain and shear strain image data for the region in which at least one portion ofthe vessel is located may be generated using speckle tracking. In one or more embodiments, the speckle tracking may include using multi—dimensional correlation of sampled echo data of one or more speckle regions (e.g., windows) undergoing deformation in the region in which the at least one portion of a blood vessel is located. The multi-dimensional ation may include ining a cross-correlation peak for the sampled pulse-echo data based on phase and magnitude gradients ofthe cross-correlated sampled pulse—echo data, PCT/U52012/033584 Such generation of strain and shear strain image data allows for the identiﬁcation of at least one vascular characteristic ofthe region in which at least one portion of a blood vessel is located (block 36) (e.g., a ﬂow characteristic associated with ﬂow through the blood vessel, a structural characteristic ated with the blood vessel, a hemodynamic characteristic associated with the blood vessel, etc). One or more types of strain and shear strain image data may be used and/or ized to identify such vascular characteristics.

Strain calculations include performing band-limited gradient calculations on the 2D (or 3D) displacement ﬁelds obtained using speckle tracking (in 2D or 3D). ime speckle tracking on full image sizes (i.e., for every pixel of the RF echo data) may be achievable in the axial direction. 2D and 3D speckle tracking may be achieved in real time in a region of interest around the blood vessel. The strains and shear s can be overlaid on B-mode or other ound imaging formats. For e, information regarding such ations are also found in Liu et al., “Real-Time 2-D Temperature Imaging Using Ultrasound,” IEEE Transactions on Biomedical Engineering, vol. 57, no. 1, pp. 12—1 6 (Jan. 2010), which is hereby incorporated by reference herein.

For example, generation of strain and shear strain image data for the region in which the at least one portion ofthe vessel is located using e tracking may include generation of at least one of axial strain and/or axial shear strain image data (e.g., axial relating to the axis through the vessel being imaged). Further, such generation of strain and shear strain image data may include generation of lateral strain and/or lateral shear strain image data. Such data may be visualized, for example, in both longitudinal views (e.g., along the blood vessel) or cross-section views (e.g., orthogonal to the axis ofthe blood vessel), as ﬁthher provided herein.

For example, one or more types of strain and shear strain image data may be used and/or visualized to identify one or more vessel wall boundaries, including the vessel wall boundaries around the entire blood vessel (e.g., such boundaries may be visualized in cross-section and/or measured about the entire blood vessel). One or more other vascular characteristics may be ﬁed, measured or calculated therefrom, such as tissue property within the one or more vessel wall ries, one or more portions of a plaque architecture adjacent the one or more vessel wall boundaries, and/or one or more PCT/U52012/033584 hemodynamic measurements based on both the motion tracking ofthe vessel wall and the blood ﬂow simultaneously.

Certain types of strain and shear strain image data may be more beneﬁcial for imaging one or more portions ofthe blood vessel and ﬂow therethrough than others. For example, axial strain image data may be beneﬁcial in identifying a ﬁrst set of opposing wall boundaries (e.g., on opposite sides of the vessel) while the axial shear strain image data may be beneﬁcial in identifying a second set of opposing wall ries (e.g., on opposite sides of the vessel) such that the boundaries ofthe entire vessel (e.g., discernible in cross-section) can be identiﬁed. Further, the same is generally the case for lateral strain and lateral shear strain. Further, for example, the lateral shear strain may be beneficial in providing wall shear stress data (e.g., used for identiﬁcation of le plaque formation).

Further, for example, shear strain images ﬁirther deﬁne the vessel walls, not only the proximal and distal walls, but also the side walls, which are hard to see on conventional ultrasound. (2D/3D+Time) calculations may also be used to reﬁne the detection ofthe wall boundaries as a function of time during the heart cycle.

The speckle tracking applied to the pulse—echo data (block 34) may include using any multi-dimensional correlation of d pulse-echo data of one or more speckle regions (i.e., windows being tracked). For example, two—dimensional correlation of sampled pulse-echo data of one or more speckle s may be used, as well as other multi-dimensional correlation ques.

In one embodiment, for example, speckle tracking is performed as described in E.

S. , “Phase—coupled two—dimensional speckle tracking algorithm,” IEEE Trans.

Ultrason, Ferroelect., FI'eq. Cantu, vol. 53, no. 5, pp. 972—990, May 2006 (hereinafter “Ebbini 2006”), which is incorporated herein by reference. For e, in general, such speckle tracking includes coarsely searching the magnitude of the sampled pulse- echo data in a lateral and axial ion to locate a vicinity of the cross-correlation peak within the cross—correlated sampled pulse—echo data; determining, within the vicinity of the correlation peak, at least two opposing gradient vectors in proximity to the cross—correlation peak; determining, within the ty ofthe cross—correlation peak, a PCT/U52012/033584 zero—phase line ofthe cross—correlated sampled pulse—echo data; and using the at least two opposing nt vectors in proximity to the cross—correlation peak and the zero— phase line to estimate the correlation peak.

More speciﬁcally, as described in Ebbini (2006), a two—dimensional (2-D) speckle tracking method for displacement estimation based on the gradients ofthe magnitude and phase of2—D complex correlation in a search region is provided. This approach couples the phase and magnitude nts near the correlation peak to determine its coordinates with subsample accuracy in both axial and lateral directions. This is achieved with a minimum level of lateral interpolation determined from the angles between the ude and phase gradient vectors on the sampled (laterally interpolated) 2-D correlation grid. One result behind this algorithm is that the magnitude gradient vectors’ ﬁnal approach to the true peak is orthogonal to the zero—phase contour.

This leads to a 2-D robust projection on the zero—phase contour that results in subsample accuracy at interpolation levels well below those needed. r, the approach includes a robust fast search thm that allows the localization of the true peak without the need for exhaustive search.

In other words, the speckle tracking uses the phase of the 2-D complex cross correlation for robust and efﬁcient estimation of displacement from speckle data. This speckle tracking method ﬁnds the true peak ofthe 2—D complex cross correlation as a constrained optimization problem. The ive of this optimization problem is to ﬁnd the coordinates ofthe axial and l lags at the true peak ofthe 2—D complex cross correlation subject to the zero~phase constraint. The basis for this formulation is shown mathematically from the inverse of the Fourier transform of the 2—D cross spectrum near the true correlation peak. This geometric approach ﬁnds both the axial and lateral displacement estimates with subsample accuracy. The method is based on the fact that the gradient vectors ofthe magnitude ofthe 2*D cross ation ch the true peak along the orthogonal to the zero—phase contour. Knowing that the zero~phase contour also passes through the true peak, it is possible to locate this peak simply by ﬁnding the point on this r at which the magnitude gradient vectors are orthogonal, provided these s ate from a grid point that is sufﬁciently close to the peak. One feature of this algorithm is that interpolation ofthe complex cross correlation is used at a minimum level that allows for a valid projection to be made. Therefore, the algorithm is —36— PCT/U52012/033584 computationally efﬁcient in terms of determining the level of olation needed for subsample accuracy from the properties ofthe underlying 2—D, cross—correlation function, not from the desired lateral resolution.

The following is a mathematical basis for correlation-based 2—D speckle tracking and relates it to a 2-D cross spectrum approach. Thereafter, two implementations of a phase-coupled 2—D speckle ng algorithm are then provided.

Let s(x, z, to) be the analytic ultrasonic signal ed from a 2-D region (e.g., speckle region or window) at time to with spatial coordinates x and 2 representing the l and axial directions, respectively. The received signal model assumes a linear space-invariant imaging system with rectangular sampling (e.g., linear array). Aﬁer undergoing translation dx and d; (respectively in the x and 2 directions), the received signal at time, t], 5(x, 2, 1]) == s(x - a}, z - dz, [0) has a 2—D Fourier transform: (kx,kz,t1) :3 SW“kg?tow—J(ismdab-+iczdg,,)a (1) where In[ and k; are the l ﬁequency variables (in rad/m) in the x and 2 directions, respectively. The 2—D cross spectrum is given by: F12U’v’m [92) II 51km, kt1)5*(lﬁm 162,160) a ligament?attests“ [130} Motivated by this equation, one investigator developed an algorithm based on the 2—D cross um for estimation of ﬁne displacement in both the axial and lateral directions. An iterative weighted least squares approach to estimate the slopes of the axial and l frequency components has been proposed. However, a more efﬁcient on to this problem can be obtained by ﬁnding the true peak ofthe 2-D cross— ation function, given by (3) and (4), PCT/U52012/033584 1 m 00 was, a). = 4W2 f 112031 anaemia:“WE-warning ~00 - -OO/ 1 90 .00 . ‘ a “ —0c) 4—00 I {"1102} “ (11'qu ~— dz); (4-) where [X and l; are the lags in the x and 2 directions, respectively. This function peaks at lag values equal to the shift values in both the axial and lateral directions. This result drives all correlation-based, 2-D displacement tracking methods. However, with the use of sampled correlation functions due to the te nature ofthe RP data collection, the true correlation peak must be found with subsample accuracy. This is especially true for the lateral displacement estimation in which the sampling interval (spacing n A— lines) is on the order of 10 times the axial ng interval.

Insight into this problem can be ed by evaluating the cross—correlation function in the vicinity of the true peak where kx(lx — dx) 3 0 and kz(lz - dz) : 0, from (3): 1 m /w lSC‘km: 13241301) [Eli +Jlk9:(l:z: “, . 7.12 (£33,. £3) % (lat) 4W2 / + Axdz ~w d3)])c‘lfﬂxdfﬁz (5)1 ‘ : 7'11“): 0) +jmiaszam " dm) “l‘ Alexa: ”" (33)]: Where: 1 DO 00 4"V f |S(km,kx,»tollzkmdkmdkg,, ~00 / -00 <6) 1 ‘00 GO Mm: . o ,2 f [swamkz,t0)|“k:gdkmdkz. (fr). 4' 5T“ ~00./-o<:: PCT/U52012/033584 [137} The factor MJG is d to the mean lateral frequency ofthe 2—D auto spectrum. rly, the factor MEX is related to the mean axial frequency ofthe 2—D auto spectrum.

Due to the fact that the axial component in many coherent imaging systems has a carrier and that the complex envelope is used, {My} :3) 0. This modulation property does not exist for the lateral factor Mm and its magnitude is generally small. It is important to note, however, that in a speckle environment, [M15] 99 O (typically lMgl 33> [szly The result in (5) can be seen as a lization of a 1—D e tracking case in which the slope ofthe phase curve is equal to the center ﬁequency ofthe (analytic) echo signal [8].

In fact, expressing (4) in vector form (and including the quadratic phase term in the Taylor series expansion ofthe exponential), we get: ’l"1a(1)=-%§W/-:f"raring-tweak3 k a}) 9531‘(Idildk s /an:to))1(1+jk’(lwde7X} éﬂdll — dXJJ2Jall: . n m na, a —— E(1 __ dx) [M __ thﬁ) (1— (1):) W.;] + j [M Mm] (1w ax)’ (9). where the elements Wxx, sz, Wm, WZZ are given by: PCT/U52012/033584 1 DC? (1351) Hi = 4,.2 /N / [SlkmEchtgjlzkidkmdkm ”(X3 ' _m (10) 1 00 DO . ‘ ' Wm = 1.90%,k3,tg)|2isgfdkmdkz, 4W3 /“ . -m mf In) Him: 2 ﬁlm; 1 0‘3 PO . ,. . t—co ~ . wot: Note that Wxx is related to the lateral frequency bandwidth of the imaging system.

It is easy to 866 that y12(dx, dz) = m0, 0) > 702(1):, la W1» 1:) 7": dz: 619- Also note that , dz) is real, which implies that the true peak of the 2~D cross ation must lie on the ase contour deﬁned by: Zﬁf‘igamazy) % ﬂif 3% - (ix) + lid—331a}; __ dz), [ﬁrm him] (1- dx) : m’ﬂ — (IX); (13) However, based on the above approximation, the magnitude ofthe 2D cross correlation is given by: W120“ I V” {711(010) ~— (1* dx)"W(1 —— that»2 +(111'(1 — dx))2j {14) for which the ose of the gradient vector can be evaluated as shown in (15), which indicates that the gradient is identically zero at the true peak (1 = dx) as expected. 617120)! —2 («(1,0) — (1 - dx)'W(l .. dx» (1_dx),w + (111,0 _ dx)) m, = (15) \/(')‘11(0, 0) — (1— dx)’W(l — dx))3 + (111’(l , CRUZ PCT/U52012/033584 Furthermore, the magnitude of the magnitude gradient near 1 = dx is proportional to the distance from the true peak, i.e., the closer the grid point to the true peak, the smaller the magnitude of the gradient. In addition, ifthe grid point is such that l — dx is orthogonal to m (m ’(l — dx) = 0), then the magnitude gradient is orthogonal to m, and the true peak can be obtained by ﬁnding the intersection ofthe line along the magnitude gradient with the tangent to the zero phase contour. Of course, this is only valid when the grid point is sufﬁciently close to the true peak. This can be determined from the ude of the ude gradient on the grid. This result provides us with insight for approach to determine the true peak of the 2—D cross-correlation function ﬁom a grid of computed values around the peak. Speciﬁcally, by comparing the slope ofthe minimum magnitude gradient with the slope ofthe phase contour in the Vicinity of the peak, one can ine whether the true peak is sufﬁciently close to the grid points to make a valid approximation.

This leads to algorithms as described herein for ﬁnding the true peak that are both robust and numerically efﬁcient in the sense that interpolation is used at a minimum level to ensure the ions of a valid projection onto the zero—phase contour. The formulation bed herein also provides the basis for an zation procedure of ﬁnding a point on a quadratic surface (magnitude of the 2-D cross correlation) subject to the zero—phase condition. This problem can be solved using the Lagrange multiplier method as described herein.

The above result can be easily extended to the case of afﬁne transformation [13]: 3(X1,t1) = 5(X0(X1:)3t0)3 (16,} where: X0: = Txl ~— th (17') (3:33- :750‘ (3511: 1 +523] 3 (1'8) 63113 r—w-ﬁNHMy» M l 1 + 3mm[ [:33] w [dz] which accounts for translation (dx), strain (em- and egg), and shear strain (ex,— and ea).

PCT/U52012/033584 In this case, we use the Fourier transform of the received signal at time A in vector form: arm) = 3(1)f k, taleﬂ jﬁka. ldx) /}T1,, . :19)I where T’ is the ose of T and I - ] is the determinant of a matrix. Note that (19) is just a generalization ofthe scaled property ofthe r transform in 1—D, which has previously been used in the analysis of decorrelation asonic echoes in the presence of strains. It is also consistent with the 2—D/3-D ation shown in [5], which analyzes the combined effects of ation and waveform warping on the variance of the displacement estimate.

It suggests that both axial and lateral strains can affect the displacement estimates based on phase matching alone. It also may provide an opportunity to directly (or iteratively) estimate the strains on the tissue region interrogated by the . We note here that, in cases in which the strain parameters, em ex, ex, and ex, have non-negligible values, the zero-phase contour does not necessarily go through the true peak ofthe magnitude ofthe 2-D cross correlation. To illustrate this point, (19) can be written explicitly in terms of the shift and strain parameters as shown in (20), which shows that the scaling ofthe Fourier transform in the kx and k: nates is coupled through the shear strain parameters. Amplitude scaling leads to decorrelation effects previously reported for 1~D tracking. Furthermore, the ted shift from the zero—phase contour is also coupled, i.e., a shift in the x direction contributes to the estimated shift in the z direction and vice versa (through the shear strain ters). The simple result given by (5) is only approximately valid for inﬁnitesimal s within the tracking window.

Fortunately, (20) provides a method to detect when the strain effects Within the window are signiﬁcant enough to affect the phase estimation. “emzka: + (i + ex.z)kz 3(1‘33151‘721“) _ 11S ((1 +€zz)}'3:c “ 5::ka to) ' IT lTl ’ lTl 7’ ‘i' 3:3)d3: "’ szdz)kx + ((1 “l" 6:7:m)d3 __C:E'C§Q£SZ) (20) ~e (((1 ”7| {157] PCT/U52012/033584 In the speckle tracking method, one is interested in the use ofphase—coupled approach in ﬁnding the true peak ofthe 2-D cross—correlation function when the tissue is undergoing cement and/or inﬁnitesimal strains. This can be considered a ﬁrst step in estimating the displacement vectors needed in applications such as, for example, vector velocity estimation and elastography. Such applications may require displacement ng based on the peak ofthe 2-D cross ation as a common step, but they may differ in the deformation model parameters used. This can be done by implementing a post-processing step that extracts the deformation ters from the shift estimation results ed from ﬁnding the true peak of the 2-D cross correlation.

Such post-processing depends on the speciﬁc problem to be determined.

An Exemplary Phase~C0upled 2-D, Speckle Tracking Algorithm The steps ofthe ary speckle tracking algorithm are guided by the properties ofthe 2—D correlation of the speckle region undergoing motion and/or deformation. The magnitude of the 2—D correlation with the complex envelope has a well behaved peak with extent in axial and lateral directions proportional to the speckle cell size. The main idea is to use a fast search algorithm to find a correlation-based match within a search Window (i.e., correlation values above a threshold). Ifthe current search point is far from the correlation peak, a fast search can be used with axial and lateral steps on the order of one-halfthe extent ofthe correlation cell size in the axial and lateral directions.

This allows the search to coarsely cover large regions without missing the true peak (if it exists). Once a match is found, the magnitude gradient of the cross correlation is used to find the peak cross correlation on the sampling grid (determined by the RF sampling frequency and spacing between the Aalines in the image). This is done by following a gradient ascent trajectory, which requires two to three correlation values to be computed for every point along the tory. nt search for the peak is very robust and efﬁcient from any direction if search point is within the width of the ation peak in the axial and lateral ions, e.g., within 3 dB from the peak. Once the grid-based correlation peak is found, a 3 X 3 ation grid is calculated (centered at the peak). In the immediate vicinity of the correlation peak, the true peak can be determined from the magnitude nt vectors and the zero-phase contour passing the true peak. From the 3 X 3 grid in the original sampling coordinates, a ﬁne estimate of the true correlation PCT/U52012/033584 peak is produced by coupling the phase of the 2—D cross correlation to the amplitude (gradients) in a way that will allow us to use (13).

Two exemplary ches for obtaining the ﬁne estimates are described in the following subsections.

A. Two—Dimensional Projection on the hase Contour This method interpolates the 3 X 3 grid near the true peak in the lateral lag direction by a sufficiently large factor (i.e., as small as possible) in order to estimate a magnitude gradient vector onal to the zero—phase contour. The intersection ofthe magnitude gradient and the zerophase contour is the estimated true peak of the 2—D cross correlation.

The steps ofthis algorithm are outlined as follows.

Step 0: From the 2-D autocorrelation function at to, estimate the search step size L; and 13I in the axial and lateral directions, ly. Typically, L: > 1LI due to the ﬁner sampling in the axial direction.

Step 1: Perform any fast search algorithm to ﬁnd the vicinity of correlation peak (above a threshold) using L;r and L: in a deﬁned search region.

Step 2: Without any interpolation, compute the local magnitude gradient and move along the gradient ascent trajectory. This step stops when the peak point on the uninterpolated grid is d.

Step 3: Once the maximum point on the uninterpolated grid is reached, compute cross—correlation values on 3 X 3 grid centered at the maximum. Interpolate the 3 X 3 grid laterally by a small factor (e.g., 8) and ﬁnd the line ons for the two gradient vectors closest to the true peak, but pointing in te lateral directions. If this condition is not satisfied with the t interpolation factor, increase olation by two and repeat the test. If maximum interpolation factor is reached (e.g., 128) and ~44.

PCT/U52012/033584 correlation at the olated peak is below a threshold (e.g., 0.75) displacement estimate is declared invalid and assigned a value ofNaN. Otherwise, proceed to Step 4.

Step 4: Find the line equations for the hase line ofthe ation function on the olated grid. The estimated true correlation peak is the point of intersection between the zero—phase line and the orthogonal line that passes through the point of ection ofthe two maximum—slope lines from the magnitude gradient.

For illustration purposes, the search trajectory in Fig. 6A is based on L, = 1 and L,— = 5 (from Step 0). Fig. 6A shows contours of the 2-D cross correlation with the search path shown [starting from (0,0) lag as indicated by the . The it indicates correlation lags tested by the fast-search algorithm and the - indicates lags tested by the gradient—ascent algorithm (L. = 1 and L3 = 5). Note that the true correlation peak has a well behaved set of contours that distinguish it from several false peaks in the search region shown. rmore, the 0.7 contour extends by approximately 4 lateral lags and axial lags (which justiﬁes the choice ofL. and L5). Step 1 begins at lag (0,0) and tests the cross—correlation coefﬁcient. If below the set threshold (0.65 in this case), it moves to the next point as indicated by the arrow (-1 l and -5 axial). If still below threshold, it tests correlation values along each pixel on the (predeﬁned) trajectory. In this case, the predeﬁned search trajectory is along rectangular counterclockwise loops (dimensions 2 LI rk z'+ 1 2 r» L: 51¢ i + 1 where i = 0, 1, . . . [max is the loop number).

The parameter 1max deﬁnes the extent ofthe allowed search region to ﬁnd a valid peak (Imax = 6 for the search region shown in Fig. 6A).

In this search example, the eight lags on the ﬁrst loop are tested, and no candidate peak is found (i.e., no correlation value > 0.65). The thm jumps from the last point in the ﬁrst loop (—l lateral and 0 axial) to the ﬁrst point in the second loop (—2 lateral and ~10 axial) and moves counter clockwise along the lateral lag direction. This step stops once the threshold test is successﬁil, i.e., correlation value at the current correlation lag above the chosen threshold. In Fig. 6A, this step stops at lateral lag 0 and axial lag ~10 after a threshold of 0.65 was reached. Step 2 also is illustrated with the help of Fig. 6A. The - signs at (0, 10), (1, 10)(l, 11), (l, 12), (2, 12), (2, 13), (2, 14) are the grid points tested by the gradient ascent. The neighboring points of (2, ~ l4) also are tested before declaring this point as the maximum point on the correlation grid.

WO 42455 PCT/U52012/033584 The ﬁnal two steps ofthe algorithm may be rated with the help of Fig. 6B which shows the magnitude and phase contours ofthe interpolated 2~D cross correlation near the true peak. Further, Figure 6B shows magnitude and phase contours ofthe 2—D cross correlation on the laterally—interpolated 3 X 3 grid in the vicinity ofthe correlation peak (between lags — 13 and —15 axially and 1 and 3 laterally). The arrows ent the magnitude gradient vectors on the interpolated grid. Lateral interpolation by a factor of 16 is used in this case (with the interpolated grid points indicated by the arrow bases).

The phase contours are labeled with phase values in radians and appear to be almost straight with a small tilt. The true peak is indicated by the open circle on the zero—phase line. The dash—dotted lines are the directions ofthe magnitude gradient vectors closest to the peak and the tangent and the orthogonal to the zero—phase line. In addition, one can see the ude gradient vectors ng in the (general) direction of the true peak.

An interpolation factor of 16 was used for this case between lateral lags 1 and 3 and axial lags -13 and ~15 (i.e., centered at 2 lateral and —14 axial from Step 2). Four (thick dash~dotted) lines are drawn along the two closest magnitude gradient vectors, the tangent to the zero—phase line, and the orthogonal to the zero—phase passing through the point of intersection of the maximum—slope lines. The true correlation peak, indicated by the open , is the point of intersection ofthe latter with the zero—phase line.

The ﬁnal step in the projection on zero—phase algorithm is an approximation ofthe magnitude gradient vector orthogonal to the zero-phase contour at the true peak. The approximation error depends on the level of olation used and can be controlled by using a lateral interpolation factor just enough to ensure that at least one of the two magnitude—gradient vectors used in the ﬁnal approximation is not substantially parallel to the zero—phase contour near the true peak. In l, this conditicn is needed only when the axial shift is practically equal to integer multiple of the axial sampling interval.

For example, in Fig. 6B one can see the ude gradient vectors at axial lag —14 approaching from the right are almost parallel to the zero-phase line. The interpolation factor of 16 used here was just enough to produce a valid projection. A valid projection is one in which the two magnitude gradient vectors closest to the peak intersect the t ofthe zero—phase line at points in which the tangent approximates the phase contour well. The key point here, however, is that the olation can be performed adaptively, thus minimizing any ssary calculations in the cement estimation PCT/U52012/033584 algorithm. In addition to the computational age of this ch, one can reduce or eliminate the error due to interpolation.

H7“ It should be noted that one or more different steps may be implemented in the method, and the speciﬁc steps ed herein are not take to be limiting on the disclosure. For example, the exemplary search algorithm may be implemented using one or more other l signal processing approaches. Further, for example, for template matching, one can use the autocorrelation sequence in the vicinity of the (0,0) lag. A related issue is the ation ofan invalid estimate based on the ation threshold in Step 3 and assigning a value ofNaN. This can be thought of as a ﬂag to be used in subsequent processing if necessary. For example, one may use an adaptive window size to maximize the composite signal-to—noise ratio (SNR). Alternatively, one may use an explicit ation model similar to that described by (19). p73 Further, the present disclosure is not limited to use of 2D speckle tracking. For e, other multi—dimensional tracking methods, such as 3D speckle tracking methods may also be applicable in one or more embodiments herein. uvq B. Surface Polynomial Fit with Phase Constraints pvn The 2—D phase projection approach described above can be thought of as an efficient method for implementation of an optimization procedure for ﬁnding the true peak ofthe complex 2-D cross correlation. The peak ﬁnding problem can be cast as ﬁnding the ients of a polynomial ﬁt to the surface of the magnitude of the 2-D cross correlation in the vicinity ofthe true peak. The nominal and interpolated contour plots shown in Figures 6A and 6B are obtained from actual imaging data and are representative of what can be expected from a standard imaging scanner. It is quite clear that the e near the true correlation peak is well behaved and appears to be quadratic in the Ix, [2 space. A polynomial ﬁt to this (smooth) surface is given by: Qilx la") 2 013:2 + big-2 ‘i’ CAKES + {are + 5Z2 + f: u7m (21) PCT/U52012/033584 [mm where [I and I; represent lateral and axial lags, respectively, and q is the magnitude of the 2—D, x cross ation in the vicinity of the true peak. It is possible to solve for the polynomial coefficients by minimizing the square error: Jtﬂ) = E: (amaze) —- ing, <22) [mm i=1 [mm where 6 = [a, b, c, d, 3, ﬂ” and N is number ofpoints on the grid (in this case N = um] This amounts to solving the over-determined system of equations: um] a=ae mm ' [W12(£mlﬁlgl)l~ [:312 3'312 Earl [31 Earl Zgl 1 b hilgaifzzlznjl Z322 11:2 53321-752 £312 £32 1 - d {34), * lfl'ﬂizxwwlz‘n'ﬂ. , g 5 lam; IEN leZr-‘rrv' [SEN ZEN 1 6 [mm This results in: Jim = (61 ~ My <61 — A9} <25“) [mu By taking the gradient of this quadratic function with respect to the real vector 6 and equating to zero, we obtain the solution: a : {A’Afl A”, [mm (2?) [mm which is valid when the matrix A is well conditioned. If this is not the case, then a regularized solution (e.g., using singular value decomposition) is sought [17]. Once the PCT/U52012/033584 coefﬁcients 6 are obtained, it is a simple matter to obtain [max and [max by analytically evaluating the gradients: 36 e = elm +. . 2bi” + 6,, . (29) and solving the matrix equation: [tarmac] — uc 20; C "d ’ __ 21)] [—6] . (30} However, this solution may amount to a form of interpolation based on magnitude only, which may result in unacceptably high levels of bias and variance in the estimation of the true correlation peak, especially in the axial direction. This is due to the almost ﬂat nature of the surface described by q in the axial direction (axial extent of only i] lag samples). However, q typically has a distinct peak in the l direction, and the derivative in this direction yields a reliable estimate ofthe peak. It is known that the analytic nature of ultrasonic echo signal allows for high subsample accuracy in axial shift estimation by using x cross correlation without interpolation. An improved solution may be ed ifthe lateral estimate from surface ﬁtting solution described by (3 O) with the axial estimate obtained from the analytic RF data.

Accounting for phase information in 2-D can be done easily by using (13) as a constraint in ction with an appropriate cost function (e.g., using the Lagrange multiplier method). To do this, we recognize that the polynomial ﬁt ﬁinction given in (21) is itself a quadratic cost function in lx and 1:. That is, one can rewrite (21) in the form: PCT/U52012/033584 title — 523:5) lz —— dag) : l’RL l (31)- Where dxo and dzo are the coordinates ofthe center ofthe 3 X 3 grid in the ty ofthe true peak and l = [Ix-dxo, ] ’. The elements ofR are obtained ﬁom the solution to least squares problem in (23). This can be shown by expanding the translated vector form in (31) and comparing with the coefﬁcients of 12x, [25 and 1x15 in (21). The optimization problem is to ﬁnd the ple shift vector 51: [5h 55] ’ (note that lmax = [dxo, dzo] ’+[6x, 65] ’) that minimizes (31) subject to the aints: ”1‘9? : — m ’m = 1, which implies The ﬁrst constraint isjust a normalization that is typically used in eigenvalue problems. The second constraint restricts the solution vector to be orthogonal to the zerophase line at the true peak (13). The elements ofm are obtained from the center frequency values in (6) and (7).

The constraints above allow us to deﬁne a new cost function in terms of the Lagrange multipliers, A and ,u: J (51) = 51’1151— At51’51)+n5l’mi (‘32) The solution to this problem can be obtained by taking the gradient ofJ with respect to 61: BJ i , ' ——- = ‘3’R51 —— 2Ab] + inn I 3 . (3 g) PCT/U52012/033584 Multiplying by m’ and g for ,u: {1, H —~2111’R61 -- ZAm’61 (34) H —2111’R51, (35) where the orthogonality constraint was used to drop the second term in (34).

Substituting ,u back in (33), we obtain the eigenvalue problem: 213m -— 2Ac‘il —-— 2111’Ra1 : 0 (36'). :f» (I - 111111’) 1151 = Adlt The solution to this eigenvalue problem is the eigenvector ated with the maximum eigenvalue of the matrix (I vmm ’)R. The Lagrange multiplier, it, is the eigenvalue, which appropriately scales the eigenvector to give the true lag (at the true maximum). The projection matrix, (1 ~-mm’), serves to align the solution obtained based on the magnitude—only method to be orthogonal to the ed slope of the zero—phase contour. Thus the ng between the phase and magnitude characteristics for ﬁnding the true peak of the 2~D cross correlation is te.

Various steps, routines, or processes may be implemented with the exemplary speckle tracking methods described herein (such as those bed in Ebbini (2006)), that further assist in ing useful image data or enhancement of image data. For example, when performing speckle tracking of speckle regions in the region in which at least one portion of a blood vessel is located, if the Speckle region (i.e., window or speckle cell) being tracked is partially within the vessel wall deﬁning the at least one portion of the blood vessel and partially outside ofthe vessel wall (e.g., lying lly in the blood within the vessel or lying partially outside ofthe boundaries ofthe vessel wall), the speckle tracking s may be lt to carry out or result in data that is inaccurate (e.g., difﬁcult to cross correlate, etc). As such, at least in one embodiment, once a vascular characteristic is identiﬁed or determined (e.g., the identiﬁcation of one or more boundaries (or portions thereof) of the vessel based on speckle tracking), the vascular characteristic may be used in performing the speckle tracking method.

PCT/U52012/033584 For e, one or more vessel wall boundaries may be identiﬁed based on the speckle ng of one or more speckle regions. Once such boundaries are identiﬁed, they can be used in the speckle tracking process. In one or more embodiments, for example, a characteristic of the one or more speckle regions (e.g., location, size, shape, etc.) may be modiﬁed based on the at least one identiﬁed vascular characteristic (e.g., such as the identiﬁcation of vessel wall boundaries). For example, the location of at least one of the one or more speckle regions being tracked may be modiﬁed (e.g., or any other teristic of one or more speckle regions, such as size, or shape, may be modiﬁed) based on the one or more vessel wall boundaries identiﬁed (e.g., such that the speckle region being tracked is entirely within or outside ofthe vessel wall). In other words, ifthe speckle tracking s determines that a e region or window being tracked is partially within the vessel wall deﬁning the at least one portion of the blood vessel and partially e of the vessel wall (e.g., lying partially in the blood within the vessel or lying partially outside ofthe boundaries of the vessel wall), then the speckle region location may be modiﬁed or otherwise adjusted such the speckle region is entirely located within the vessel wall or outside of the vessel wall. Further, the size or shape (e.g., narrowness or width or ) ofthe speckle region tracked may be modiﬁed during the speckle tracking such that the speckle region is entirely located within the vessel wall or outside ofthe vessel wall, or provides y estimates, e.g., of displacement. In other words the speckle region is modiﬁed such that it falls entirely outside the vessel wall, entirely within the vessel wall, or entirely within the blood, based on the one or more boundaries determined for the vessel (e.g., by prior speckle ng and generation of strain and/or shear strain image data for the region in which the at least one p01tion of the vessel is located).

For example, generally, ng windows are designed to optimize or achieve a tradeoff between spatial resolution (e.g., generally improved by reducing the window size) and reducing the variance ofthe displacement estimates (e.g., generally, by increasing the window size). For example, in one embodiment, for a uniform speckle region, the signal—to—noise (SNR) of the echo data, together with the transducer bandwidth, are the primary s in selecting the window size. rmore, in the absence of severe deformation, windows are designed to be approximately square, i.e., approximately the same axial and lateral ions. Near the vessel wall, especially on WO 42455 PCT/U52012/033584 the lumen side, many assumptions may be violated. High frame rate imaging (e.g., M2D mode imaging) is provided to mitigate some ofthese effects, but there may also be a need to account directly for deformations within the tracking window (e.g., part ofthe window being in the vessel and part in the blood leading to deformation issues). In such cases, one may apply adaptive window size and shape design or selection algorithms that will produce optimal estimates (e.g., best possible displacement estimate). The 2D phase—coupled algorithm provides feedback on the quality of the estimates (e.g., ﬁgures ofmerit). These ﬁgures t may be used in terizing the quality ofthe estimate obtained, for example, using different windows with the same height, but different Width and vice versa. As such, it is possible to run these windows in parallel and implement a voting scheme that selects the estimates with the highest ﬁgures of merit and reject or weigh down (e.g., apply a lower weighting factor) the estimates with lower ﬁgures of merit.

Still further, vessel wall echo reverberations may corrupt pulse—echo data received from the blood. As such, in one or more embodiments of speckle ng of one or more speckle regions within the blood, the speckle tracking may remove echo components in the pulse—echo data clue to reﬂection at the vessel wall when performing such speckle ng ofthe pulse-echo data ﬁom the one or more speckle s in the blood (e. g., using a dereverberation ﬁlter).

In other words, as described herein, various e tracking methods (e.g., two— ional speckle tracking methods) can be used to image tissue motion and deformations in the vicinity ofblood vessels (e.g., for use in detecting and staging of vascular e). However, in one or more embodiments, vessel wall echo reverberations may overwhelm the echoes (scattering) from the blood and result in loss ofﬂow information in large regions within the vessel. One embodiment that may be used to correct for such reverberations includes use ofa time-varying dereverberation inverse ﬁlter for echo data within the vessel (e.g., echo data from the blood). The design for such a ﬁlter may be inﬂuenced by the fact that the reverberation pattern varies signiﬁcantly with the pulsatory motion ofthe vessel wall. Minute changes in the on/orientation ofthe vessel wall with respect to the imaging beam result in measurable changes in the e—specular echo mixture at the vessel wall and the observed periodicities in the reverberation pattern Within the vessel. Therefore, a time— PCT/U52012/033584 varying inverse ﬁlter may be used to remove the reverberation components appropriately during the heart cycle. e tracking (e. g., two-dimensional e tracking) can be used for the analysis oftissue motion and deformation, especially in vascular imaging applications.

In addition to the direct measurements and/or characterization ofthe vessel cs, such speckle tracking may provide tissue cement ﬁelds and underlying anatomical information le for important challenge areas such as computational ﬂuid dynamics.

For example, as described herein, simultaneous estimation of tissue motion/deformation and blood flow vector ty in a blood vessel (e.g., the human carotid artery) using pulse—echo diagnostic ultrasound can be med (e.g., using phase~coupled two—dimensional speckle tracking which achieves subsample displacement estimation cy in both axial and lateral directions together with a high frame rate (e.g., 325 frames per second) M2D g mode both lateral ﬂow velocity and wall motion can be estimated simultaneously thus providing pressure (through vessel diameter) and ﬂow (through vector velocity». Such measurements can provide the basis for hemodynamic computations that may be used in the assessment of vessel wall compliance, an important indicator of the health of the vessel as described herein. r, as described herein, axial and l displacement ﬁelds are well-behaved and allow for strain and shear strain calculations in both tissue and blood. Together with anatomical image information, these ty/strain ﬁelds may provide input for computational ﬂuid dynamic models, which may allow for inverse calculations suitable for the assessment of the health of the vasculature and surrounding tissue (e.g. detection and staging of atherosclerosis).

One possible limitation oftwo-dimensional speckle tracking s based on pulse—echo ultrasound is the reverberation components associated with strong specular reﬂectors. Reverberation produces clutter signal components that mix with back scattering components from regions distal to the specular reﬂector. This is signiﬁcant in blood vessels where the scattering from blood is typically 30 ~ 40 dB below the specular reﬂection from the vessel wall. For applications where both the tissue motion and blood velocity vectors are impmtant, however, the r due to reverberation from the region PCT/U52012/033584 within the vessel walls should be minimized. An algorithm for dereverberation of pulse— echo ultrasound data to restore echoes from the blood scattering region prior to two— dimensional speckle tracking may be used to address such reverberations. Due to the changes in the scattering characteristics ofthe vessel wall during the different phases of the heart cycle, it is observed that the reverberation signal components are non— nary, which indicates that a time—varying inverse ﬁlter for dereverberation may be needed. One possible filter design approach may utilize short—range correlation for the echo signal from the wall and the long-range ation of the echo signal from the al wall and the vessel region. Periodicities in the correlation functions are attributed to the vessel wall architecture and multilayer tissue ure resulting in distinct specular components (eg. adventia).

The following provides one or more implementations of dereverberation or examples testing the same. Such speciﬁc implementations or examples are not to be taken as being limiting to the present disclosure.

Exemplary Dereverberation Filtering Information Data Acquisition for the example Dereverberation Filtering A Sonix RP (Ultrasonix, Canada) ultrasound scanner loaded with a program used for MZD pulse—echo data collection. Collected data is then lined to a controller PC through Gigabit Ethernet for real-time data processing. The data sing computer can easily handle the intensive computations ed by high resolution (both spatial and temporal) speckle ng and separable 2D post ﬁltering by utilizing a many core GPU (nVlDIA, Santa Clara, CA) A linear array probe (LAl4-5/3 8) was used to acquire all data of this example. The center cy of the transmit pulse on the probe was 7.5 MHZ.

Received Signal Model A received signal model for echo beamformed ultrasound data obtained ﬁom a typical scanner and typical vascular probe such as those bed above is provided.

The echo data forming one image line is given by PCT/U52012/033584 x,.(r,l) 2237431) +x,:(r,l) (l) where xcﬁ‘; 1) represents the coherent echo ent from depth 2 =c/t in image line I and x,{t; Z) is the incoherent echo component ﬁ'om the same location. At the vessel wall, the coherent ent is largely due to specular reﬂection from the multi—layered vessel and supporting structure. For simplicity, assume a single layer in a homogeneous medium and a narrowband model: 27 m 22 21%]; KAI-J) : Rer (I ”" “39) (2). . ‘l‘ X 06k}? (I " "B ‘“ ' C C k=l u) r ’ 6 t' __ T ( )is) where [3 (t) ‘“‘ («Hawk the analytic transmitted ultrasound pulse, RN is the reﬂection coefﬁcient, and dw is the vessel wall thickness. In (2), Rm, ents the tissue-wall reﬂection coefﬁcient and ak is a ﬁmction of the wall—tissue reﬂection and ission coefﬁcients. The reﬂection coefﬁcients are typically small (e.g. < 10%) and the series in (2) is practically 2 to 4 terms for each layer. Unfortunately, the reverberation terms interfere with the echo components from the blood in the .

Despite their rapid decay, their ude remains high enough to mask the echoes ﬁom the blood in a region that extends several millimeters inside the vessel. A dereverberation ﬁlter can be used to unmask the echoes from the blood and allow the mensional speckle tracking to estimate vector velocity inside the vessel.

Inverse Filter Design for Echo Dereverberation The correlation function of the coherent echo component exhibits secondary peaks at rk =2kdw/c. The amplitude ofthese peaks diminishes exponentially with k (due to multiple reﬂection within the wall). These secondary peaks can be estimated from the autocorrelation function ofthe echo data after baseband conversion. Results from actual vascular imaging experiments suggest that a Gaussian mixture model (GMM) may be most appropriate for ng the probability density function (pdt) ofthe clean echo data. The GMM is also motivated by a number of eses on the scattering by blood (e.g. due to ﬂow or red blood cell aggregation). An inﬁnite impulse response (HR) model is assumed for the dereverberation inverse ﬁlter PCT/U52012/033584 ﬁn] :: ﬁn] — Z 611;)?[11 - k] (3) [(521 The coefﬁcients, {Gk }k=l can be ed by maximizing the log—likelihood with respect to the GMMs in the ﬂow channel: [2323 where Ng is the number of Gaussian densities in the GMIVI (each represented by {wg/Ji, 01}) and M, is the length of output test sequence. The parameters of the HR inverse ﬁlter can be obtained by maximizing 3’ in (4) by partial tives with respect to {aklii—Til, 8$__1 "0+er Ng ‘3 (33[n] -,LL;)3*[12——k] Z“ M- ()5 861k :N) (52 {233] 17—-2110 i: i where é-‘Eaand ”5"»: are the ith term inside the inner summation in (4) and the inner summation, respectively. The parameters ofthe ﬁlter can be obtained iteratively through (33’ (1/;[171 + l] = (1k [/72] + 5W (6) where m is the ion index and d is chosen sufﬁciently small to allow for ﬁne convergence (at the expense of convergence speed). In at least one embodiment, (5) can be simpliﬁed ifthe GMIM is dominated by one distribution, which is often the case. At any rate, an empirical values of 5 =0:01 result in convergence within 10 iterations or less. order Selection: The parameter N in (3) can be determined from the identiﬁcation of secondary peaks in the autocorrelation ﬁmction of the complex envelope PCT/U52012/033584 ofthe echo data ﬁom the proximal vessel wall and from the vessel including the proximal wall. Periodicities in these correlation functions are ﬁable with the wall (and supporting-tissue) architecture, which is primarily responsible for reverberation within the . For the carotid artery of an typical human subject, the order varies depending on the phase of the cardiac cycle with a range of 20 — 45 (lower orders when the vessel wall is maximally stretched). A maximum value, Nmax on the model order could be set from the observation of the icities throughout the heart cycle.

Reverberation-free Training Data ing reverberation—free echo data can be accomplished by tilting the imaging transducer to avoid specular ions from the vessel wall (after identifying the vessel wall). In practice, a very small tilt is sufﬁcient to signiﬁcantly remove the specular echo and maintain a good view ofthe vessel. Color ﬂow or power Doppler mode can be used to deﬁne the ﬂow region for extracting the training echo data. Once this is achieved, the imaging transducer is returned to a position where the vessel ries can be best deﬁned from specular reﬂections. It is also possible to get reverberation-free data ﬁom corrupted views by carefully selecting s where the reverberation components are not present.

For example, a blood vessel with 6—mm diameter may provide 2 mm region of reverberation-free echo data at its distal end. One can use the latter approach to identify the GMIVI during a typical heart cycle. Based on actual echo data from a human subject, the value chg in (4) is typically 3 to 4, all with zero mean. arying Filter Design Algorithm Based on the above derivation and facts about the data acquisition model for beamformed echo data from typical vascular imaging ultrasound, the following ﬁlter design steps are provided. For image line 1 and image frame p, Step 0 Detect the proximal vessel wall from complex envelope data. ~58— PCT/U52012/033584 Step 1 Determine the HR model order, N, from the short range and long—range correlation functions.

Step 2 Determine the GlVEM parameters from the reverberation free segment of the vessel echo data.

Step 3 Determine the ﬁlter coefﬁcients using the m likelihood algorithm (Equations 4 — 6).

Step 4 Apply the HR ﬁlter to vessel data with appropriate initial conditions to avoid breaks in the echo data.

Step 5 Compute long—range correlation function on y[n]. If reverberation still detectable, While N < Nmax GoTo Step 3. Otherwise Stop. s and Discussion regarding Example: Data Collection — The Sonix RP was used to collect MZD mode data from imaging the right carotid artery of a y volunteer. Real-time beamformed RF data was collected at a frame rate of 325 fps and processed off—line to produce images similar to those shown in s 16A—16B, which shows lateral (Figure 16A) and axial cements (Figure 168) ofthe carotid artery coded (in color or gray scale) and overlaid on the B-mode gray scale . The displacement fields were estimated using the phase coupled two—dimensional speckle tracking algorithm described in Ebbini 2006.

Approximately 1.9 seconds (642 frames) of data was acquired capturing more than three heart cycles to allow for modeling the dynamics of echoes and reverberation ﬁ‘om the vessel, primarily due to the al wall. Spatio—temporal maps ofthe echoes from the vessel (proximal wall through the vessel, but excluding the distal wall) are shown in Figure 17A. The image shows the envelope of the RF data from one image line (approximately at the center ofthe images). This is ed to as M—mode ultrasound and it allows for the analysis of wall motion in cardiac and vascular imaging application. _59_ PCT/U52012/033584 Figure 17B shows an M—mode image ofthe echoes from the proximal wall only during the acquisition period. The image illustrates the dynamics of the wall echoes during the heart cycle. For example, for the ﬁrst few lines (0 < t < 0: ls), the wall represents a purely specular reﬂector. On the other hand, the wall echo appears less zed when the wall undergoes ion due to pressure changes in the vessel.

Examining the envelope in the vessel shown in Figure 17A, one can see a strong reverberation pattern down to about 4 mm into the vessel. This pattern masks any scattering from the blood and makes it difﬁcult, ifnot impossible, to estimate the ﬂow velocity vectors in a large region within the vessel.

Figures ISA-18B show spatio—temporal maps ofnormalized correlation ﬁmction of the complex envelope from the vessel including the proximal wall (Figure 18A — a long range correlation) and proximal wall only (Figures 18B —— a short range correlation) corresponding to the M—mode images shown in Figures l7A-l7B. One can see the strong secondary peaks at z 3: 0:8 mm when the vessel wall is not stretched and s as a specular reﬂector. On the other hand, this peak is diminished and the main lobe of the correlation function is ned when the vessel wall is stretched. This result demonstrates that a time—varying dereverberation ﬁlter will be ial. [254} Figures l9A—l9D illustrate the performance ofthe dereverberation ﬁlters obtained through the proposed maximum hood algorithm. Using short—range and long—range correlation data, the filter order was determined to be 21 for two frames (Frame 1 at t = 0.003 s and Frame 100 at 1‘: 0.31 sec). The GMM models were estimated from the vessel data near the distal wall using Ng = 3. Ny = 50 was used in both cases and 5 = 0:005. Convergence was achieved in 7 and 9 ions for Frame 1 and Frame 100 data, respectively. The magnitude response of both ﬁlters demonstrate e ﬁltering and dereverberation characteristics; signal bandwidth 2 — 6 MHZ is inverted and ripple throughout. Figures 19A-19D represent the dereverberation ﬁlters and vessel data with and without filtering for Filter Frame 1 (Figures l9A-19B) and Filter Frame 100 (Figures D) (the ﬁltered signals generally being closer to 0 line axis than the al).

Results from vascular imaging experiment from a healthy eer demonstrate the feasibility of dereverberation of echo data ﬁom blood vessels using time—varying inverse rberation ﬁlters to account for the time—varying nature ofthe reverberation ~60— 2012/033584 components during the heart cycle. Such ﬁltering may for more accurate vector velocity estimation in the vessel using speckle ng (e.g., two—dimensional e tracking) methods. The exemplary algorithm is robust and computationally efﬁcient and requires minimal training making it well-suited for real—time ultrasound imaging applications.

The following provides one or more implementations ofvascular imaging and characterization ofvessel wall dynamics or examples testing the same, such as generally described herein. Such speciﬁc implementations or examples are not to be taken as being ng to the present disclosure.

Generally, a method for simultaneous imaging oftissue motion and ﬂow with subsample accuracy in both axial and lateral directions is illustrated. The method utilizes a phase-coupled 2D e tracking approach, which employs the true 2D complex cross correlation to ﬁnd sub—pixel displacements in both axial and l directions. The imaging sequence on a Sonix RP scanner has been d to allow high frame rate 2D data collection in a limited ﬁeld ofview covering the region of interest (MZD—mode). er with the robust 2D e tracking method, M2D imaging allows for capturing the full dynamics of the ﬂow and wall/tissue motion, even when the ﬂow is primarily in the lateral direction (with respect to the g beam).

The ﬁne vector displacement tes in both axial and lateral directions are shown to allow for smooth and contiguous strain and shear strain calculations with minimal ﬁltering. The quality ofthe displacement and strain ﬁelds is demonstrated by experimental results from a ﬂow phantom (ATS Model 524) and in viva images of the carotid artery in a healthy volunteer. The results demonstrate simultaneous imaging of the vector ﬂow ﬁeld and the wall/tissue motion and the corresponding strains at high spatial and temporal sampling. This may e an essential tool in ng the ﬂuid— solid interactions between the blood and blood vessel.

Materials andMethodsfor an Exemplary Imaging Method Coupled 2D Speckle Tracking — The phase—coupled 2D speckle tracking algorithm used is described in Ebbini 2006 and also at least partially described herein.

The speckle tracking method is based on the gradients of the magnitude and phase of2D complex correlation in a search region. This approach couples the phase and magnitude PCT/U52012/033584 gradients near the correlation peak to determine its coordinates with subsample accuracy in both axial and lateral directions. This is achieved with a m level of lateral interpolation determined from the angles between the magnitude and phase gradient vectors on the d (laterally interpolated) 2D cross—correlation grid. One result behind this algorithm is that the magnitude gradient vectors” final approach to the true peak is orthogonal to the zero—phase contour. This leads to a 2D robust projection on the hase contour that s in subsample accuracy at interpolation levels well below those needed using previously proposed methods. It is shown that estimated 2D vector displacement ﬁelds obtained using the phase—coupled technique display a full range of values covering the c range without evidence of quantization. In comparison, a previously published method using 1D phase—projection after lateral interpolation produces severely quantized lateral displacement fields (at the same levels of interpolation as the 2D, phase—coupled method).

Data Acquisition - A Sonix RP (Ultrasonix, Canada) ultrasound scanner loaded with a program used for high frame—rate MZD pulse-echo data collection is used.

Collected data is then streamlined to a controller PC through t Ethernet for real— time data processing. The data processing computer can easily handle the ive computations required by high resolution (both spatial and al) e tracking and separable 2D post ﬁltering by utilizing a many core GPU (nVlDIA, Santa Clara, CA). A linear array probe (LA14—5/3 8) was used to acquire all data shown in this paper.

The center frequency of the transmit pulse on the probe was 7.5 MHZ.

Experiments using the ATS Model 524 flow phantom and a Cole—Farmer MasterFlex roller pump were conducted to illustrate the displacement tracking in axial and lateral directions. M2D data was collected using the Sonix RP scanner at a frame rate of 111 fps. Images of a 4—mm flow channel in the ATS phantom were collected under controlled fluid flow with an appropriate speed setting of the MasterFlex pump to mimic typical blood ﬂow rates (6.g. 336 ml/min) in carotid arteries. Cellulose pheres were diluted in water to produce linear ring ﬁom the ﬂow channel during data collection (pure water was also imaged as a l to determine the channel boundaries as a ground truth). ~62— PCT/U52012/033584 Strain ation — Strain and shear strain calculations were performed using MATLAB’S nt command on the axial and lateral displacement ﬁelds obtained using the 2D speckle ng algorithm. The strain and shear strain computations are followed by a simple Gaussian lowpass ﬁlter with size of 3 3 and standard deviation of l, i.e. minimal post ﬁltering of the strain and shear strain ﬁelds. s and Discussion Regarding the Exemplary Imaging Method Experimental Phantom Result The 4-mm flow channel was imaged using the LAl4—5 probe with the channel axis perpendicular to the beam axis, i.e. lateral flow. Typical examples of the resulting strain and shear strain ﬁelds are shown in Figures 8A—8B and Figures 9A—9B. Figures 8A~8B Show axial strain and axial shear strain ofthe 4mm ﬂow channel longitudinal walls, respectively. Figures 9A-9B show lateral strain and lateral shear strain of the 4—mm ﬂow channel longitudinal walls, respectively. One can see the smoothness ofthe strain ﬁelds which demonstrates the well—behaved nature ofthe displacement ﬁelds. At the same time, one can appreciate the clear ation between the channel and the surrounding tissue-mimicking material.

The dynamics of the estimated vector ﬂow ﬁeld are rated by the result in Figure l0 which shows a plot of the periodic channel diameter (dashed lines) (obtained from the axial component at the channel walls computed ﬁom the tracked channel wall displacement) and the corresponding average lateral ﬂow velocity in the channel (solid lines) for several cycles ofpump operation (i.e., within the 4 mm flow channel in the ATS phantom). The result shows clearly the quasi—periodic nature of the observed ﬂow ty and the phase relationship n the diameter ure in the channel) and flow. Note that the small negative ccmponent in the ﬂow velocity occurs right after the diameter reaches its maximum value (i.e. minimum re in the channel). This “back up” of the ﬂuid in the channel was easily observed in B-Inode movies, but the wall motion was much more subtle (< 150mm) and was only seen in the axial displacement ﬁeld.

WO 42455 PCT/U52012/033584 The dynamic behavior of the lateral ﬂow along the axis ofthe imaging beam (at a lateral distance of —3 .2 mm) is illustrated by Figure ll. One can see the , well— behaved nature ofthe lateral displacement ﬁelds consistent with the quasi-periodic pattern shown in Figure 10. It is noted that the results shown in Figure ll are minimally processed, i.e., the continuity and the high SNR of the vector displacement estimation in the ﬂuid and the surrounding tissue is a direct consequence ofthe proper application of the phase—coupled 2D speckle tracking method.

In Vivo Experiment A t ofthe d artery of a healthy volunteer was imaged using the LAM— probe at 1 ll fps. As in the phantom case, the axial and lateral displacement ﬁelds were contiguous throughout the region of interest and allowed for the ation ofthe strain and shear strain ﬁelds shown in Figures 12A—12B and Figures l3A-lBB. Figures 12A~12B Show axial strain and axial shear strain ofthe carotid artery longitudinal vessel walls, respectively. Figures l3A—13B show l strain and lateral shear strain of the carotid artery longitudinal vessel walls, respectively. One can see the clear demarcation between the vessel and the wall/perivascular tissue in the axial strain and lateral shear strain images. Also, the pulsation effect can be appreciated from the axial shear strain and lateral strain images. These results demonstrate reliable tion ofthe vector motion fields (axial and lateral) and their utility in ing realistic strain and shear strain ﬁelds.

Finally, we show a cross sectional View ofthe strain ﬁelds around the carotid artery of a y volunteer. The strains and shear strains in both axial and lateral directions are shown in Figures l4A~l4B and s ISA-15B, respectively. Figures l4A—l4B show axial strain and axial shear strain ofthe carotid artery cross-sectional vessel walls, respectively. Figures ISA—15B show lateral strain and lateral shear strain ofthe carotid artery sectional vessel walls, respectively. Examining Figures 14A-14B one can observe that the motion in the axial direction produced a strain in the axial direction at the wall (Figure 14A) and a shear component to the lateral direction (Figure 14B). The direction ofthe strains in this case depicts an expanding vessel. Similarly, the motion in the lateral direction produced a strain in the lateral direction (Figure 15A) and a shear component to the axial direction as in (Figure 15B). Movies of the strain and shear PCT/U52012/033584 strain ﬁelds have showed the dynamics of wall movement clearly (such not table in this application). In addition to the wall dynamics, the strain images provide a tool for identifying the boundaries of the vessel in both axial and lateral directions. The latter is generally difﬁcult to determine from B-mode images.

As a result of the example, a method for imaging the cement and strain ﬁelds in the vicinity of ﬂow channels has been demonstrated experimentally in a ﬂow phantom and in viva imaging ofthe carotid artery ofhealthy volunteer. The results show that, at sufﬁciently high frame rates, speckle tracking methods produce ehaved displacement estimates of both the tissue motion and ﬂow in the channel. These displacement ﬁelds are well suited for strain and shear strain calculations with minimum ﬁltering. Further, it has been demonstrated that time rms ofﬂow velocity and pressure to follow the periodic motion ofthe roller pump in the phantom experiments.

Furthermore, the axial wall displacements (indicative of pressure) and average lateral ﬂow velocity in the channel have a clear phase relationship. This indicates that the method used may be useful in obtaining the full dynamic motion ﬁelds suitable for use in solid-ﬂuid interface ng of vascular mechanics.

All patents, patent documents, and references cited herein are orated in their entirety as if each were incorporated separately. This disclosure has been provided with reference to illustrative embodiments and is not meant to be construed in a limiting sense. As described previously, one skilled in the art will recognize that other various rative applications may use the techniques as described herein to take advantage of the beneﬁcial characteristics ofthe apparatus and methods described . Various modiﬁcations of the illustrative embodiments, as well as onal embodiments ofthe disclosure, will be apparent upon reference to this description. —65-

Claims

1. An g method comprising: providing ultrasound pulse-echo data of a region in which at least one portion of a blood vessel is located, wherein the pulse-echo data comprises pulse-echo data at a frame rate such that measured cement of the vessel wall defining the at least one portion of the blood vessel and measured average blood flow through the at least one portion of the blood vessel have a quasi-periodic profile over time to allow motion tracking of both the vessel wall and the blood flow simultaneously; generating strain and shear strain image data for the region in which the at least one n of the vessel is located using speckle tracking, wherein the speckle tracking comprises using multi-dimensional correlation of pulse-echo data of one or more speckle regions undergoing deformation in the region in which the at least one portion of a blood vessel is located, n the dimensional correlation comprises determining a crosscorrelation peak for the sampled pulse-echo data based on phase and magnitude nts of the cross-correlated pulse-echo data; and identifying at least one vascular characteristic of the region in which at least one portion of a blood vessel is located based on the strain and shear strain image data, wherein the at least one vascular characteristic comprises at least one of a flow characteristic associated with flow through the blood vessel, a structural characteristic associated with the blood vessel, and a hemodynamic characteristic associated with the blood vessel.

2. The method of claim 1, wherein identifying at least one vascular characteristic comprises identifying one or more vessel wall boundaries.

3. The method of claim 1 or claim 2, wherein the speckle ng further comprises ing a characteristic of at least one of the one or more speckle s being d based on the one or more vessel wall boundaries identified such that the at least one speckle region is entirely within or outside of the vessel wall.

4. The method of any of claims 1 -3, wherein identifying one or more vessel wall boundaries comprises identifying vessel wall boundaries around the entire blood vessel.

5. The method of any of claims 1-4, wherein identifying at least one vascular characteristic ses measuring tissue property within the one or more vessel wall ries.

6. The method of any of claims 1-5, wherein identifying at least one vascular characteristic comprises identifying one or more portions of a plaque architecture adjacent the one or more vessel wall boundaries.

7. The method of any of claims 1-6, n identifying at least one vascular characteristic comprises calculating one or more hemodynamic measurements based on both the motion tracking motion of the vessel wall and the blood flow simultaneously.

8. The method of any of claims 1-7, wherein using multi-dimensional correlation of sampled pulse-echo data of one or more speckle regions comprises using mensional correlation of sampled pulse-echo data of one or more speckle s.

9. The method of any one of claims 1-8, wherein ining the cross-correlation peak ses: coarsely searching the magnitude of the sampled pulse-echo data in a lateral and axial direction to locate a vicinity of the cross-correlation peak within the cross-correlated sampled pulse-echo data; determining, within the vicinity of the cross-correlation peak, at least two opposing gradient vectors in proximity to the cross-correlation peak; determining, within the vicinity of the cross-correlation peak, a hase line of the cross-correlated sampled pulse-echo data; and using the at least two opposing gradient vectors in proximity to the cross-correlation peak and the zero-phase line to estimate the cross-correlation peak.

10. The method of any of claims 1-9, wherein the method further comprises determining the usage of ultrasonic energy based on the identification of the at least one vascular characteristic of the region in which at least one portion of a blood vessel is located.

11. The method of claim 10 , wherein the method further comprises providing at least one transducer configured to it and e ultrasonic energy, wherein the at least one transducer is used to obtain the pulse-echo data and to generate ultrasonic energy.

12. The method of any of claims 1-11, wherein generating strain and shear strain image data for the region in which the at least one portion of the vessel is located using ensional e tracking comprises generating at least one of axial strain and axial shear strain image data and/or lateral strain and lateral shear strain image data.

13 The method of any of claims 1-12, wherein providing ultrasound pulse-echo data of a region in which at least one portion of a blood vessel is located comprises using coded tion.

14. The method of any of claims 1-13, wherein the method further comprises applying a dereverberation filter to the pulse-echo data from one or more speckle regions in the blood to remove echo components in the pulse-echo data due to tion at the vessel wall when performing speckle tracking of the pulse-echo data from the one or more speckle regions in the blood.

15. A system for vascular imaging, comprising: one or more ultrasound transducers, wherein the one or more transducers are configured to deliver ultrasound energy to a vascular region resulting in pulse-echo data rom; and means for controlling the capture of pulse-echo data at a frame rate such that ed displacement of a vessel wall defining at least one portion of a blood vessel in the vascular region and measured average blood flow through the at least one portion of the blood vessel have a quasi-periodic profile over time to allow motion tracking of both the vessel wall and the blood flow simultaneously; means for generating strain and shear strain image data for the region in which the at least one portion of the vessel is located using speckle tracking, wherein the speckle tracking comprises using dimensional correlation of pulse-echo data of one or more speckle regions undergoing ation in the region in which the at least one portion of a blood vessel is located, wherein the multi-dimensional correlation comprises determining a cross- correlation peak for the sampled pulse-echo data based on phase and magnitude gradients of the cross-correlated pulse-echo data; and means for identifying at least one vascular characteristic of the vascular region in which at least one portion of a blood vessel is located based on the strain and shear strain image data, wherein the at least one vascular characteristic comprises at least one of a flow characteristic associated with flow through the blood vessel, a structural characteristic associated with the blood vessel, and a hemodynamic characteristic associated with the blood vessel.

16. A system for vascular imaging ing to claim 15, comprising: one or more ultrasound transducers, wherein the one or more transducers are configured to deliver ultrasound energy to a vascular region resulting in pulse-echo data therefrom; and processing apparatus configured to: control the capture of pulse-echo data at a frame rate such that measured displacement of a vessel wall defining at least one portion of a blood vessel in the vascular region and measured average blood flow through the at least one portion of the blood vessel have a quasi-periodic profile over time to allow motion ng of both the vessel wall and the blood flow simultaneously; generate strain and shear strain image data for the region in which the at least one portion of the vessel is located using speckle tracking, wherein the speckle tracking comprises using multi-dimensional correlation of pulse-echo data of one or more speckle regions oing ation in the region in which the at least one portion of a blood vessel is located, wherein the multi-dimensional correlation comprises determining a crosscorrelation peak for the sampled pulse-echo data based on phase and magnitude gradients of the cross-correlated pulse-echo data; and identify at least one vascular characteristic of the vascular region in which at least one portion of a blood vessel is located based on the strain and shear strain image data, n the at least one vascular characteristic ses at least one of a flow characteristic associated with flow through the blood vessel, a ural characteristic associated with the blood , and a hemodynamic characteristic associated with the blood vessel.

17. A system for vascular imaging ing to claim 16, comprising: one or more ultrasound transducers, wherein the one or more transducers are configured to deliver ultrasound energy to a vascular region resulting in pulse-echo data rom; and sing apparatus configured to: control the capture of pulse-echo data of the vascular region in which at least one portion of a blood vessel is located; use speckle ng of one or more speckle regions of the ar region in which at least one portion of the blood vessel is located to track motion of both the vessel wall defining the at least one portion of the blood vessel and the blood flow through the at least one portion of the blood vessel, wherein the pulse-echo data is captured at a frame rate such that displacement of the vessel wall defining the at least one portion of the blood vessel and blood flow through the at least one portion of the blood vessel are measurable aneously within a same periodic cycle corresponding to a cardiac pulse cycle; and identify at least one vascular characteristic of the vascular region in which the at least one portion of the blood vessel is located based on the simultaneously measured displacement of the vessel wall and average blood flow.

18. The system of claim 17 , wherein the processing apparatus is further operable to generate strain and shear strain image data for the region in which the at least one n of the vessel is located using the speckle tracking, wherein the speckle tracking comprises using multi-dimensional correlation of sampled pulse-echo data of the one or more speckle regions undergoing deformation in the region in which the at least one portion of a blood vessel is located, wherein the multi-dimensional correlation ses determining a cross-correlation peak for the sampled pulse-echo data based on phase and magnitude gradients of the orrelated sampled pulse-echo data.

19. The system of any of claims 16 to 18, wherein the processing apparatus is operable to identify at least one ar characteristic comprises identifying one or more vessel wall boundaries.

20. The system of claim 19, n the processing apparatus is operable, when using the speckle tracking, to modify a characteristic of at least one of the one or more speckle regions being tracked based on the one or more vessel wall boundaries identified such that the at least one speckle region is entirely within or outside of the vessel wall.

21. The system of any of claims 19-20, wherein the processing apparatus is operable to identify vessel wall boundaries around the entire blood vessel.

22. The system of any of claims 19-21, wherein the processing apparatus is operable to measure tissue property within the one or more vessel wall ries.

23. The system of any of claims 19-22, wherein the processing apparatus is le to identify one or more portions of a plaque architecture adjacent the one or more vessel wall boundaries.

24. The system of any of claims 19-23, wherein the processing apparatus is operable to calculate one or more hemodynamic measurements based on both the motion tracking motion of the vessel wall and the blood flow simultaneously.

25. The system of any of claims 19-24, wherein the processing apparatus is operable to use mensional correlation of sampled pulse-echo data of one or more speckle regions.

26. The system of any of claims 19-25, n the processing apparatus is operable to control the determination of the cross-correlation peak by at least: coarsely searching the ude of the sampled pulse-echo data in a lateral and axial direction to locate a vicinity of the cross-correlation peak within the cross-correlated sampled pulse-echo data; determining, within the vicinity of the cross-correlation peak, at least two opposing gradient vectors in proximity to the cross-correlation peak; determining, within the vicinity of the cross-correlation peak, a hase line of the correlated sampled pulse-echo data; and using the at least two opposing gradient vectors in ity to the cross-correlation peak and the hase line to estimate the cross-correlation peak.

27. The system of any of claims 19-26, wherein system further includes a therapy system to deliver therapy to a patient based on the identification of the at least one vascular characteristic of the region in which at least one portion of a blood vessel is located.

28. The system of claim 27, wherein the therapy system comprises a system operable to use ultrasonic energy to deliver therapy based on the identification of the at least one vascular characteristic of the region in which at least one portion of a blood vessel is located.

29. The system of any of claims 27-28, wherein the therapy apparatus comprises at least one transducer configured to transmit and receive ultrasonic energy, wherein the at least one transducer is operable to provide ultrasonic energy to deliver therapy based on the identification of the at least one ar characteristic of the region in which at least one portion of a blood vessel is d and the at least one transducer is operable for use in obtaining the pulse-echo data to generate image data.

30. The system of any of claims 16-29, wherein the sing apparatus is operable to generate strain and shear strain image data for the region in which the at least one portion of the vessel is located using two-dimensional speckle tracking, wherein using two-dimensional speckle tracking comprises generating at least one of axial strain and axial shear strain image data and/or lateral strain and l shear strain image data.

31. The system of any of claims 16-30, wherein the sing apparatus is operable to control ing ultrasound pulse-echo data of a region in which at least one portion of a blood vessel is located using coded excitation.

32. The system of any of claims 16-30, wherein the processing apparatus is le to apply a dereverberation filter to the pulse-echo data from one or more speckle regions in the blood to remove echo components in the pulse-echo data due to reflection at the vessel wall when performing speckle tracking of the echo data from the one or more speckle regions in the blood.

33. The system of claim 15, wherein the system further includes therapy means for delivering y to a patient based on the identification of the at least one vascular characteristic of the region in which at least one portion of a blood vessel is located.

34. The system of claim 33, wherein the therapy means comprises a system operable to use ultrasonic energy to deliver y based on the identification of the at least one vascular characteristic of the region in which at least one portion of a blood vessel is located.

35. The system of claim 15, claim 33 or claim 34, wherein the therapy means comprises at least one transducer configured to transmit and receive onic energy, wherein the at least one transducer is operable to provide ultrasonic energy to r therapy based on the identification of the at least one vascular characteristic of the region in which at least one portion of a blood vessel is located and the at least one transducer is operable for use in obtaining the pulse-echo data to generate image data. PCT/U