Chen, Shigao

Elasticity imaging has emerged as a viable clinical tool for assisting in diagnosis of certain diseases such as liver fibrosis and cancer. Many methods have been developed for generating and measuring shear waves in soft tissues. The advantage of using shear waves is that the shear wave velocity is proportional to the mechanical material properties of the tissues under investigation. Soft tissues are inherently viscoelastic and considerable effort has been made to quantify the dispersion, or variation of frequency, of the velocity and attenuation of shear waves. This is accomplished by measuring the shear wave motion and using Fourier-based techniques to extract the shear wave velocity and attenuation. Additional considerations are made in tissues where geometric dispersion is also present such as in the heart. Viscoelastic characterization of the shear viscoelasticity of soft tissues in vivo such as human liver, human kidney, and swine heart will be shown. We will also demonstrate parameterization of the results by using a model-free approach or by fitting the shear wave velocity dispersion to rheological models. The diagnostic value of the viscoelastic parameters will be discussed for each particular application. [This work was supported in part by NIH Grant Nos. DK092255, DK082408, and EB002167.]

We validate a technique for model-free measurement of shear wave velocity and attenuation. A mechanical shaker was used to excite harmonic plane and cylindrical waves in phantoms and excised tissue. Radiation force was used to excite impulsive cylindrical waves. 2D FFT of the displacement yielded the k-space whose coordinates are frequency and the wave number. The shear wave velocity at each frequency was obtained by finding the maximum at the given frequency in k-space and dividing the frequency coordinate by the wave number coordinate. The attenuation (α) at a given frequency was calculated using α = FWHM × π/√3, where FWHM is the full width at half maximum of the k-space peak along the given frequency. This method was applied to measure shear wave velocity and attenuation of transplanted kidneys and livers, and in the thyroid tumor, and compare it to the healthy tissues. The velocities and attenuations at each frequency for various excitation methods agree within one standard deviation. The k-space estimates of velocity and attenuation agreed with those obtained using the phase gradient (velocity) and amplitude decay (attenuation). The transplanted organs and the thyroid tumor had higher velocity and lower attenuation than healthy tissues.

Measurements of viscoelasticity with shear wave velocity dispersion requires measurements over a large bandwidth. In this study, we explored the parameters that modulate the frequency characteristics of shear waves induced using radiation force push beams. We used a Verasonics ultrasound scanner equipped with a linear array transducer. We performed measurements of shear wave motion induced using both focused and unfocused ultrasound beams. Measurements were made in elastic phantoms with shear moduli of 1, 4, and 16 kPa. The number of elements used for the unfocused beams were varied from 8 to 24, and for the focused beams from 16 to 128. The shear wave motion was tracked using plane wave imaging, and a one-dimensional autocorrelation algorithm applied to the acquired in-phase/quadrature data. At each pixel we calculated the fast Fourier transform of the data and found the center frequency, center-of-gravity, and −3 dB bandwidth. We compared the frequency characteristics from the different push beams. The frequency characteristics were found to be spatially variant and dependent on the number of elements used as well as the shear modulus of the medium. The center frequency, center-of-gravity, and the bandwidth were found to be correlated to one another, and strongly associated with the stiffness of the medium.

Robust shear-wave motion detection is essential for producing reliable shear elasticity measurements for ultrasound shear-wave elastography (SWE). However, because shear-wave motion is extracted from radiofrequency signals which can be heavily contaminated by noise, shear-wave motion detection can be very challenging, especially for in vivo applications. This study investigated the implementation of harmonic imaging (HI) to facilitate more robust shear-wave tracking based on HI’s effectiveness in suppressing imaging noise associated with ultrasound reverberation, phase aberration, and clutter noise. A HI shear-wave tracking sequence was developed combining the pulse-inversion HI method with the plane wave imaging technique to transmit phase-inverted pulses at a high frame rate of several kilohertz. The backscattered ultrasound signals from phase-inverted pulses were added to suppress the fundamental and enhance the second harmonic component, from which the shear-wave motion was extracted. A pork belly phantom experiment showed that HI could significantly improve shear-wave motion detection by producing almost three-fold less underestimation of shear-wave motion and over ten-fold more precision for shear-wave speed measurements than fundamental imaging. An in vivo transthoracic case study of a human heart showed that HI substantially improved the success rate of shear-wave motion detection and could provide consistent estimates of the left ventricle myocardium stiffness.

Thyroid cancer is the fastest growing age and gender adjusted cancer in 2011 according the American Cancer Society. The majority of the clinically diagnosed thyroid nodules are benign while less than 5% represent intrathyroidal cancers. Currently, the clinical gold-standard procedure for assessing the thyroid nodules is needle biopsy, a procedure that is associated with significant financial burden as well as pain and risk for patients. Therefore, a noninvasive, affordable, and potentially widely available method to differentiate between benign and malignant thyroid nodules can play an important role in reducing the number of unnecessary biopsies. In this study, we investigate the feasibility of two acoustic radiation force elastography techniques, shear wave dispersion ultrasound vibrometry (SDUV) and comb-push ultrasound shear wave elastography (CUSE imaging), in identifying thyroid nodules (imaging) and differentiating between benign and malignant pathologies based on their elasticity and viscosity (SDUV measurements). Our preliminary results show that the measured shear elasticity and shear viscosity parameters depend on tissue type; hence, these measurements may be utilized to differentiate between healthy normal thyroid tissue, benign nodules, and malignant nodules. Further studies on a large population of patients is required to better evaluate the role of the combination of elasticity and viscosity properties of tissue in differentiating various thyroid nodules.

Vibro-acoustography is an imaging technique that maps the acoustic response of an object to a localized harmonic radiation force. This force is generated by two interfering continuous-wave ultrasound beams at slightly different frequencies f1 and f2. The system point-spread function (PSF) is related to the radiation force on a point-target. Imaging artifacts depend on the PSF sidelobes, which can be reduced by mismatching the sidelobes of each ultrasound beam. Here, we propose a beamforming approach based on an 8-element sector transducer with consecutive elements alternately driven at f1 and f2. The transducer generates two ultrasound beams skewed by 22.5° with respect to each other. The system PSF is analytically derived. The theory is validated by experiments using a small steel sphere (radius =0.2 mm) as a point-target. A laser vibrometer is used to measure the vibration of the sphere and evaluate the PSF of the system. Theoretically, the PSF sidelobes are under −15.8 dB in eight spots circularly distributed and separated by 22.5°, which agrees with the experimental results. Simulation shows that with 16 elements sidelobes are under −35.2 dB. In conclusion, sector transducers for vibro-acoustography may have lower sidelobes as the number of array elements is increased. [Work supported by Grant Nos. EB00535-01, EB2640, and IMG0100744.]

Elasticity imaging methods have been used to study tissue mechanical properties and have demonstrated that tissue elasticity changes with disease state. Quantitative mechanical properties can be measured in a model independent manner if both shear wave speed and attenuation are known. However, measuring shear wave speed attenuation is challenging in the field of elasticity imaging. Typically, only shear wave speed is measured and rheological models, such as Kelvin-Voigt, Maxwell and Standard Linear Solid, are used to solve for shear viscoelastic complex modulus. Acoustic radiation force has been used to study quasi-static viscoelastic properties of tissue during creep and relaxation conditions, however, as with shear wave propagation methods, a rheological model needs to be fit to the creep or relaxation experimental data to solve for viscoelastic parameters. This paper presents a method to quantify viscoelastic properties in a model-independent way by estimating complex shear elastic modulus over a wide frequency range using time-dependent creep response induced by acoustic radiation force. The acoustic radiation force induced creep (RFIC) method uses a conversion formula that is the analytic solution of the constitutive equation relating time dependent stress and time dependent strain. The RFIC method in combination with shear wave propagation is used to measure the complex shear modulus so that knowledge of the applied radiation force magnitude is not necessary. Numerical simulation of creep strain and compliance using the Kelvin-Voigt model shown that the conversion formula is sensitive to sampling frequency, the first reliable measure in time and the long term viscosity approximation. Experimental data are obtained in homogeneous tissue mimicking phantoms and excised swine kidneys.

This study concerns the development and preliminary experimental verification of a topological sensitivity–based platform for the material characterization of tissue anomalies exposed by vibroacoustography-type imaging techniques. Vibroacoustography (VA) is a high-resolution imaging method that has been applied to the detection of pathological changes in soft tissues. Although the data provided by this method is related to the mechanical properties of tissue, the viscoelastic parameters of the object cannot be estimated by this imaging method itself. Topological sensitivity (TS) method is a data processing methodology that can be used to estimate the viscoelastic parameters of an object from vibration data. In this study, the concept of topological sensitivity is applied to interpret the vibroacoustography measurements for the purpose of lesion characterization. In the proposed approach, the topological sensitivity function, which signifies the variation of a given cost functional when an infinitesimal inclusion with trial material parameters is placed at the location of a point force, is formulated in terms of the adjoint field. The effectiveness of the resulting formula as a material indicator for lesion characterization is demonstrated by estimating the relative elastic parameters of a well-controlled neoprene sphere embedded in a tissue-mimicking phantom specimen.

Ultrasonic contrast agents consisting of microbubbles are used to assess tissue perfusion. The microbubbles are highly reflective and nonlinear and thus produce harmonics that are stronger than those from tissues. However, the magnitude of harmonic signals resulting from a region with microbubbles also depends on the acoustic pressure of incident ultrasound and the attenuation of intervening tissues in the ultrasound path. Therefore, the harmonic magnitude, as used in traditional harmonic imaging, may not be a reliable indicator of the presence or absence of microbubbles, and hence, tissue perfusion. To compensate for these effects, we present two parameters defined as the ratio of the harmonic to the fundamental component ( HFR) and the ratio of the harmonic to squared fundamental ( HSFR). A simplified model is used to illustrate the usefulness of these two parameters. Experiments show that both parameters improve detection of microbubbles and that HSFR performs better than HFR.