Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Oct 1.
Published in final edited form as: Biomed Phys Eng Express. 2019 Oct 18;5(6):10.1088/2057-1976/ab4c05. doi: 10.1088/2057-1976/ab4c05

Attenuation Measuring Ultrasound Shearwave Elastography as a Method for Evaluating Pancreatic Viscoelasticity

Benjamin G Wood 1, Mollie E Ireson 2, Matthew W Urban 1, Ivan ZD Nenadic 2
PMCID: PMC7051008  NIHMSID: NIHMS1065012  PMID: 32123575

Abstract

Pancreatic cancer is the fourth most common cause of cancer-related fatalities as there are a limited number of tools to diagnose this disease in its early stages. Pancreatitis is characterized as an inflammation of the pancreatic tissue due to an excess amount of pancreatic enzymes remaining in the organ. Both of these diseases result in a stiffening of the tissue which makes them suitable for the use of elastography techniques as a diagnostic method. However, these methods typically assume that the tissue is purely elastic when biological tissue is inherently viscoelastic. The attenuation measuring ultrasound shear elastography (AMUSE) method, which measures both attenuation and shear wave velocity was used to characterize the viscoelasticity of pancreatic tissue. This method was tested in ex vivo normal porcine samples that were also stiffened in formalin and in vivo by conducting studies in healthy human subjects. Ex vivo testing showed ranges of phase velocity, group velocity, and phase attenuation values of 1.05 – 1.33 m/s, 0.83 – 1.12 m/s, and 183 – 210 Np/m. After immersing the ex vivo tissue in formalin there was a distinguishable difference between normal and stiffened tissue. This study produced percent difference ranges of phase velocity, group velocity, and phase attenuation from 0 to 100 minutes in formalin of 30.0% - 56.5%, 38.2% - 58.6%, and 55.8% - 64.8%, respectively. The ranges of phase velocity, group velocity, and phase attenuation results in human subjects were 1.53 – 1.60 m/s, 1.76 – 1.91 m/s, and 196 – 204 Np/m, respectively. These results were within a similar range reported by other elastography techniques. Further work with the AMUSE method in subjects with pancreatitis and cancer is needed to determine its effectiveness in showing a difference between healthy and diseased tissue in humans.

Introduction

Pancreatic cancer is the fourth most common cause of cancer-related deaths in the United States (Siegel, Miller et al. 2017). Ninety-five percent of pancreatic cancers are due to pancreatic ductal adenocarcinoma (Siegel, Miller et al. 2017). The most common physical symptoms of pancreatic cancer are jaundice, abdominal or back pain, and unintended weight loss. These symptoms rarely appear in early stages of the cancer (Chu, Goggins et al. 2017). This suggests that a method to detect the disease early by means of screening exams bi-yearly or more frequently for high risk patients would improve overall survival rates. High-risk patients with family history of cancer, particularly pancreatic cancer, or other risk factors would be identified by primary care physicians. According to the American Cancer Society, the one and five-year survival rates are 25% and 6% respectively (Siegel, Miller et al. 2017). The five-year survival rate when diagnosed in stage one is 17% (Walling and Freelove 2017), a significant improvement and sign that early detection would be highly beneficial.

Currently, pancreatic cancer is diagnosed using various imaging modalities, blood tests, and ultrasound-guided needle biopsy. Noninvasive imaging methods such as computed tomography, magnetic resonance imaging, positron emission tomography, and ultrasound imaging are used to locate cancer and help perform staging (Siegel, Miller et al. 2017). If an abnormality is detected, a biopsy of the tissue is taken for examination. If the scans fail to produce definitive results, other invasive procedures are done. Most commonly, an endoscopic ultrasound procedure is conducted, in which a thin, flexible tube (endoscope) with an ultrasound device is inserted through the mouth or nose to the area of the small intestine adjacent to the pancreas. Detailed scans of the pancreas are taken with the probe and then a small needle is used to take a biopsy sample. The results of biopsy are necessary for a definitive diagnosis (Chu, Goggins et al. 2017). At this time, there is no accurate way to noninvasively diagnose pancreatic cancer.

Pancreatic cancer, particularly pancreatic ductal adenocarcinoma (PDAC), is characterized by a dense fibrous stroma that consists of multiple cell types including pancreatic stellate cells, macrophages, and extracellular protein components such as collagen, gylcoproteins, and polysaccharides(Pelosi, Castelli et al. 2017). These constituents elevate the stiffness of the tissue. Also, PDAC has a strong antitumor immunological response, but is highly angiogenic (Pelosi, Castelli et al. 2017). While inflammatory infiltrate will typically increase fluid and tissue viscosity, the angiogenic component with added blood will influence both tissue stiffness and viscosity.

Pancreatitis is inflammation of the pancreas due to premature activation of digestive enzymes in the pancreas. Normally, the enzymes move through the pancreatic ducts to the small intestine and activate to assist in digestion. In the case of pancreatitis, the enzymes irritate the pancreas and cause inflammation. Pancreatitis can be acute or chronic, either passing after a few days or remaining multiple years, respectively (Johnson, Besselink et al. 2014, Stram, Liu et al. 2016). Pancreatitis can result in detectable changes in tissue mechanical properties in both acute and chronic pancreatitis. These changes occur due to the increase in inflammatory cells, interstitial edema, acinar cell death, and hemorrhage. The increase in fluid and other inflammatory infiltrate can affect the viscosity of the tissue at a macroscopic level and thereby increase the attenuative properties of the tissue (Silva-Vaz, Abrantes et al. 2019).

Palpation is an important part of a medical examination during which mechanical properties of a tissue are evaluated by manual compression in an effort to assess changes in tissue elasticity and viscosity. Over the past several decades, the fields of ultrasound and magnetic resonance elastography have developed various techniques to measure mechanical properties of organs including the liver, heart, breast, prostate, kidney, and arteries (Sarvazyan, Hall et al. 2011, Tanter and Fink 2014). In one class of these techniques an external vibrator or focused ultrasound is used to create propagating shear waves in the tissue, and ultrasound or magnetic resonance imaging to track the wave propagation. Various algorithms are then used to calculate the velocity of the propagated shear wave and compute tissue elasticity (Bercoff, Tanter et al. 2004, Chen, Urban et al. 2009).

These methods typically assume that the tissue is elastic. However, biological tissue is inherently viscoelastic and by ignoring the viscosity the elasticity calculations will be biased and fail to address an important component of tissue mechanics. Shear wave elasticity and viscosity are important biomechanical properties of biological tissue that are often used to characterize soft tissues (Jiang, Li et al. 2015) that can be used to understand the health of the tissue since tissue mechanical properties are affected by pathological changes. Multiple approaches have been developed to evaluate viscoelastic parameters using shear waves. Most of these methods utilize the shear wave velocity dispersion, or variation of the shear wave velocity with frequency (Chen, Urban et al. 2009, Deffieux, Montaldo et al. 2009, Rouze, Palmeri et al. 2015, Budelli, Brum et al. 2017, Rouze, Deng et al. 2017). Additional efforts have been made to measure shear wave attenuation which is also determined by the viscoelastic properties of the tissue (Rouze, Palmeri et al. 2015, Budelli, Brum et al. 2017, Nenadic, Qiang et al. 2017, Rouze, Deng et al. 2017).

Pancreatic cancer and pancreatitis alter the normal pancreatic tissue and as a result the mechanical properties. Cancer can change the viscoelastic properties of the tissue, which has been demonstrated (Tanter, Bercoff et al. 2008, Kumar, Denis et al. 2018). Inflammation processes such as those present in acute cellular rejection in liver transplants has been shown to change the phase velocity and shear wave attenuation (Nenadic, Qiang et al. 2017). Because the shear wave velocity and attenuation are sensitive to the underlying stiffness and viscosity of the tissue, we aim to develop techniques to characterize the viscoelastic properties of tissue. In this work, the attenuation measuring ultrasound shearwave elastography (AMUSE) method was used to characterize the viscoelasticity of the pancreas. This was done by utilizing the method in ex vivo porcine pancreas as a baseline. The ex vivo pancreas samples were then immersed in formalin to simulate stiffening of the tissue that would be apparent in pancreatitis and pancreatic tumors. The AMUSE method was also tested in human subjects.

Methods

Attenuation measuring ultrasound shear elastography (AMUSE)

The AMUSE method has been used previously in liver transplants for the evaluation of acute cellular rejection and livers of pigs with right heart dysfunction (Nenadic, Qiang et al. 2017, Hu, Qureshi et al. 2018). The method performs unique analysis of shear wave motion data that is detailed in Nenadic, et al. (Nenadic, Qiang et al. 2017). Briefly, a two-dimensional Fourier transform (2D-FT) is applied to the spatiotemporal representation of shear wave particle velocity v(x,t) where x is the lateral direction with respect to the ultrasound transducer along which the shear wave is propagating and t is time to obtain a the frequency-domain representation, V(k,f), as a function of spatial, k, and temporal, f, frequency. It should be noted that the spatiotemporal data needs a correction for the cylindrical wave diffraction so the amplitude is multiplied by x to account for this diffraction attenuation (Nenadic, Qiang et al. 2017). The cylindrical wave originates from the shape of the acoustic radiation force push beam utilized in this study. For each frequency of interest, f0, a profile of V(k,f0) is extracted. The peak of this profile is identified in the k-direction as k0 to evaluate the wave velocity using c(f0) = 2πk0/f0 (Bernal, Nenadic et al. 2011). The full-width at half-maximum (FWHM) is also characterized from this profile to measure the shear wave attenuation using α(f0)=πFWHM/3 (Nenadic, Qiang et al. 2017). This method was used for evaluation of the shear wave velocity and attenuation at f0 = 100 Hz. While the acoustic radiation force produces shear waves with broad frequency content, we found that 100 Hz could be used on a consistent basis for evaluating the velocity and attenuation.

Ex Vivo Pancreas Study

To evaluate the ability of AMUSE to characterize pancreatic tissue three excised porcine pancreas samples were used. For the ex vivo experiments, three porcine pancreas samples were obtained from the Department of Surgery, Mayo Clinic Rochester, Minnesota. The pigs were sacrificed by the Mayo Clinic Department of Surgery under an approved surgical training protocol assigned from the Mayo Clinic Institutional Animal Care and Use Committee using either exsanguination or Fatal Plus. Each pancreas was placed in a degassed water bath and scanned using an L7–4 linear array (Philips Healthcare, Andover, MA) connected to a Verasonics research system (V1, Verasonics, Inc., Kirkland, WA). This scanning setup is shown in Fig. 1. The pancreas was placed on a gelatin block to reduce noise and ultrasound signal scattering from the bottom of the water bath container. The samples were scanned at multiple locations as labeled in Fig. 2. Measurements were performed at each location five times. After scanning, the samples were placed in clean degassed water and frozen.

Figure 1:

Figure 1:

Open in a new tab

Ex vivo pancreas scanning setup with ultrasound linear array transducer inducing and measuring the shear waves. The pancreas was placed on a gelatin pad to reduce noise and ultrasound scattering.

Figure 2:

Figure 2:

Open in a new tab

Excised porcine pancreas with corresponding numbers for scanning positions used.

For the acquisition, a focused beam at 4.09 MHz with length of 200 μs was transmitted. The push used 64 elements and was focused at 20 mm (F/# = 1). The acoustic radiation force push was followed by detection plane wave transmissions to be used for coherent compounding. The plane waves were transmitted at three angles (−4°, 0°, +4°) for an effective frame rate of 3333 Hz (Montaldo, Tanter et al. 2009). The in-phase/quadrature (IQ) data was saved from each acquisition for offline analysis. The shear wave particle velocity was estimated using an autocorrelation method (Kasai, Namekawa et al. 1985). The particle displacement was then estimated using temporal integration. A bandpass filter with a passband of 50–600 Hz was applied to the data.

The particle displacement was used for analysis with a time-to-peak algorithm to measure the group (time-domain) shear wave velocity, and the AMUSE methodology (Palmeri, Wang et al. 2008, Nenadic, Qiang et al. 2017). The data was averaged through depths of 1.54 mm over a window centered about the focal depth of the push beam. The averaging improves the signal-to-noise ratio of the data and makes the measurements more robust (Amador, Otilio et al. 2016, Amador, Chen et al. 2017). Multiple windows of 1.54 mm in axial width were used to obtain velocity and attenuation values from a given measurement. For each averaged set of particle displacement waveforms, the AMUSE technique was applied. The group velocity measurements were accepted if the corresponding R2 value for the linear regression of time-to-peak values versus position was above 0.80. Data points for the phase velocity were accepted if the measurements were within the range of 0.7–2.0 m/s. The lower end of this range was chosen based on the values for healthy pancreas from multiple studies (Yashima, Sasahira et al. 2012, Göya, Hamidi et al. 2014, Xie, Zou et al. 2015, Zaro, Lupsor-Platon et al. 2016). The upper end was chosen as there were no data points above 2.0 m/s. The data points for phase attenuation would be accepted if the corresponding phase velocity was accepted.

Ex Vivo Formalin Study

To evaluate the specificity of AMUSE between normal and abnormal tissue, a formalin study was conducted. Formalin is a solution of formaldehyde in water that is normally used as a preservative for specimens but also causes stiffening of the tissue after continual submersion (Ling, Li et al. 2016, Nenadic, Mynderse et al. 2016). The setup for Fig. 1 was used for this study with a 10% formalin solution substituted for degassed water. The samples used were the same from the previous study but were frozen and thawed between studies. The samples were scanned at zero minutes in solution and every 20 minutes up to 100 minutes. The samples were measured eight times for each location and in the same regions as in the previous study. The acquisition details were the same as described above. The data was then analyzed with the AMUSE methodology.

In Vivo Human Subjects Study

In this proof-of-concept study, we recruited twelve human subjects for evaluation of shear wave elastography measurements in the pancreas. These subjects had no history of liver or pancreas disease. The experimental protocol was approved by the Mayo Clinic Institutional Review Board and written informed consent was obtained prior to scanning. Twelve subjects [eight females and four males, mean body mass index (BMI): 27.85 ± 5.33 kg/m2] were recruited for the study. Imaging was performed by an experienced sonographer and during breath holds. Under the guidance of conventional B-mode ultrasound imaging, the sonographer located a region of interest within the different regions of the pancreas (head, body, and tail). Scanning occurred near the subject’s midline to locate the pancreas. Five consecutive acquisitions were obtained. A Verasonics Vantage system (Verasonics, Inc., Kirkland, WA) was used with a C5–2v transducer (Verasonics, Inc., Kirkland, WA). For shear wave generation, a single focused push beam with push duration of 600 μs was transmitted. The detection beams were wide beams with an f /9.9 focal configuration transmitted with a frequency of 2 MHz. Received signals from two steering angles were compounded, giving an effective pulse repetition frequency of 2.77 kHz (Montaldo, Tanter et al. 2009). We characterized the acoustic output for the acquisitions made (Herman and Harris 2002). For the acquisitions we used settings for the Verasonics system in which the MI0.3 was less than 1.83, and the ISPTA,0.3 was less than 5.94 mW/cm2. The temperature increase measured at the transducer surface was 0.3 °C. The accepted group velocity measurements were based off the corresponding R2 being 0.80 and above. The accepted range for the phase velocity measurements was from 1.0 to 3.5 m/s. The accepted range for the phase attenuation measurements was based on the corresponding phase velocity measurements being accepted. Each subject may be represented by 1–3 data points when plotted.

Statistical Analysis

To evaluate if there were differences between the three pancreas samples, a one-way, non-parametric (Kruskal-Wallis) test using GraphPad Prism (GraphPad Software, San Diego, CA) was performed with a significance threshold of p < 0.05. This analysis was conducted on the ex vivo results across the three pancreas samples and on the in vivo results across the head, body, and tail of the pancreas. We also evaluated trends of group and phase velocity and attenuation with time while the ex vivo pancreas samples were immersed in formalin. A Pearson correlation was used to evaluate these trends for each pancreas and position. A Kruskal-Wallis test (p < 0.05) was used to analyze the phase velocity and phase attenuation results in the head, body, and tail of the pancreas from the in vivo study. The group velocity measurements were shown to be normally distributed by a Shapiro Wilkes test. Therefore, a one-way ANOVA test (p < 0.05) was used to analyze the in vivo group velocity results.

Results

Ex Vivo Pancreas Study

From this experiment, it was observed that the AMUSE method is effective in producing consistent results in the different regions of the pancreas when in an ex vivo situation. Both the individual phase velocities at 100 Hz and group velocities are reported in Fig. 3 for the corresponding sample positions. As AMUSE uses attenuation to characterize the viscoelastic parameters of tissue, the attenuation was measured at 100 Hz and reported in Fig. 4. Mean phase and group velocity values with standard deviations are reported in Table 1.

Figure 3:

Figure 3:

Open in a new tab

Phase and group velocity violin plots for ex vivo pancreas across three samples. The dark dashed line represents the median and the light dashed line represents the quartiles. The horizontal width of the violin plot at a given value serves to provide information about the distribution of the data similar to a histogram.

Figure 4:

Figure 4:

Open in a new tab

Phase attenuation violin plots for ex vivo pancreas samples. The dark dashed line represents the median and the light dashed line represents the quartiles. The horizontal width of the violin at a given value serves to provide information about the distribution of the data similar to a histogram.

Table 1:

Mean group velocity, phase velocity (100 Hz), and phase attenuation (100 Hz) with standard deviations for ex vivo pancreas samples.

Region Group Velocity, m/s Phase Velocity, m/s Attenuation, Np/m
Pancreas 1, Position 1 1.12 ± 0.25 1.33 ± 0.24 201 ± 44
Pancreas 1, Position 2 0.83 ± 0.29 1.22 ± 0.21 209 ± 85
Pancreas 1, Position 3 0.89 ± 0.26 1.29 ± 0.28 210 ± 112
Pancreas 2, Position 1 0.93 ± 0.51 1.05 ± 0.21 209 ± 76
Pancreas 2, Position 2 0.93 ± 0.45 1.16 ± 0.27 183 ± 79
Pancreas 3, Position 1 0.86 ± 0.38 1.14 ± 0.16 207 ± 69
Open in a new tab

Ex Vivo Formalin Study

To evaluate the ability of AMUSE to effectively detect a difference in measurements with stiffer tissue that would be expected of cases like pancreatitis or pancreatic cancer, formalin was used to stiffen the samples and measurements were taken every 20 minutes over 100 minutes. The change in phase and group velocities and phase attenuation over time in formalin are shown in Figs. 5(a–c), respectively. In both velocity graphs there is a consistent trend of increasing velocity with time as the pancreas was immersed in the formalin solution. Fig. 5(c) shows a continual decreasing trend as the tissue became stiffer over time. These trends are further visualized in Fig. 6, which reports the mean phase velocity, group velocity, and phase attenuation values for each individual pancreas. The mean values at zero and 100 minutes were used to calculate percent differences and standard deviation. The percent differences and standard deviation in phase velocity for pancreas one, two, and three were 56.5 ± 15.7%, 32.9 ± 13.9%, and 30.0 ± 9.12%, respectively. The percent differences and variance in group velocity for pancreas one, two, and three were 49.0 ± 32.5%, 38.2 ± 8.59%, and 58.6 ± 8.15%, respectively. The percent differences and variance in phase attenuation for pancreas one, two, and three were 63.2 ± 21.0%, 64.8 ± 18.1%, and 55.8 ± 13.0%, respectively.

Figure 5:

Figure 5:

Open in a new tab

Mean phase velocity (a), group velocity (b), and phase attenuation graphs for ex vivo pancreas in formalin solution over 100 minutes with scans starting at zero minutes and performed every 20 minutes. The black dashes over each set of data points is the median for the respective pancreas, position, and time. The legend in (c) applies to all plots.

Figure 6:

Figure 6:

Open in a new tab

(a-c) Mean phase velocity (a-c), group velocity (d-f), and phase attenuation (g-i) values with error bars for each pancreas from 0–100 minutes submerged in formalin.

Human Subjects Study

To further evaluate the feasibility of using AMUSE in the pancreas, we conducted studies in 12 human subjects. Like the previous experiments, both the phase and group velocities and the attenuation were measured in the head, body, and tail of the pancreas. The phase and group velocity measurements are reported in Fig. 7. The phase attenuation measurements are shown in Fig. 8. The mean values and standard deviations for phase velocity, group velocity, and phase attenuation are reported in Table 2.

Figure 7:

Figure 7:

Open in a new tab

Phase and group velocity graphs from human subjects in the head, body, and tail of the pancreas. The middle line represents the mean and the error bars represent the standard deviations.

Figure 8:

Figure 8:

Open in a new tab

Phase attenuation graph from human subjects in the head, body, and tail of the pancreas. The middle line represents the mean and the error bars represent the standard deviations.

Table 2:

Mean phase velocity, group velocity, and phase attenuation values with standard deviations measured in healthy human subjects.

Region Group Velocity, m/s Phase Velocity, m/s Attenuation, Np/m
Head 1.91 ± 0.33 1.53 ± 0.21 196 ± 7
Body 1.86 ± 0.31 1.60 ± 0.27 204 ± 15
Tail 1.76 ± 0.39 1.53 ± 0.17 199 ± 11
Open in a new tab

Statistical Analysis

One-way, non-parametric tests of the ex vivo data showed that the median phase velocity, group velocity, and phase attenuation values among the three pancreas samples were significantly different (p < 0.0001, p = 0.0001, p = 0.0430; Krusksal-Wallis). The results of Pearson correlation tests of the formalin study are reported in Tables 3 and 4. Table 3 shows the r-squared and p values for phase velocity, group velocity, and phase attenuation at each pancreas and position. All measurements were shown to have significant trends as all p values were below 0.05. Table 4 shows the same values for each measurement but for each organ. The table shows that all samples have significant trends. One-way, non-parametric tests of the in vivo human volunteer data showed no significant difference of the median values between the head, body, and tail of the pancreas for phase velocity and phase attenuation (p = 0.8821, p = 0.0855; Kruskal-Wallis). A one-way ANOVA test showed no significant difference of the mean group velocity values (p = 0.4981; ANOVA).

Table 3:

Pearson r and P values for phase velocity, group velocity, and phase attenuation at each pancreas and positon for the ex vivo results.

Measurement Pancreas 1 Pancreas 1 Pancreas 1 Pancreas 2 Pancreas 2 Pancreas 3
Position 1 Position 2 Position 3 Position 1 Position 2 Position 1
r P r P r P r P r P r P
Phase Velocity 0.818 <0.001 0.772 <0.001 0.668 <0.001 0.471 <0.001 0.700 <0.001 0.579 <0.001
Group Velocity 0.797 <0.001 0.393 0.022 0.357 0.011 0.374 0.0191 0.759 <0.001 0.829 <0.001
Phase Attenuation −0.806 <0.001 −0.828 <0.001 −0.488 <0.001 −0.847 <0.001 −0.746 <0.001 −0.547 <0.001
Open in a new tab

Table 4:

Pearson r and P values for phase velocity, group velocity, and phase attenuation for each individual pancreas for the ex vivo results.

Measurement Pancreas 1 Pancreas 2 Pancreas 3
r P r P r P
Phase Velocity 0.735 <0.001 0.507 <0.001 0.579 <0.001
Group Velocity 0.427 <0.001 0.421 <0.001 0.829 <0.001
Phase Attenuation −0.651 <0.001 −0.793 <0.001 −0.547 <0.001
Open in a new tab

Discussion

In this work, we explored the feasibility of using AMUSE to assess the viscoelastic properties of the pancreas in both ex vivo and in vivo situations. The ex vivo pancreas study was used to evaluate the ability of AMUSE to produce measurements in known healthy tissue without interference of surrounding body tissue. The results shown in Table 1 showed a phase velocity range of 1.02–1.33 m/s, a group velocity range of 0.82–1.12 m/s, and a phase attenuation range of 183–219 Np/m. These results show that AMUSE is capable of measurements in healthy ex vivo porcine pancreas tissue.

To evaluate the specificity of AMUSE to detect differences between normal and stiffened tissue that would be present in pancreatitis and pancreatic cancer, the ex vivo pancreas samples were immersed in formalin to simulate stiffening over time. Pearson correlation tests on the trends in Fig. 5 were reported in Tables 3 and 4. The tests showed all trends were statistically significant with the phase velocity and attenuation showing the strongest correlations. The trends of phase velocity and group velocity values over time were shown in Fig. 5(a,b). Both graphs show a general increase in velocity measurements. Pancreas one, two, and three showed phase velocity percent differences and standard deviations between 0 and 100 minutes in formalin of 56.5 ± 15.7%, 32.9 ± 13.9%, and 30.0 ± 9.12%, respectively. Group velocity percent differences and variance between 0 and 100 minutes in formalin were 49.0 ± 32.5%, 38.2 ± 8.59%, and 58.6 ± 8.15%, respectively. This is to be expected as a stiffer tissue will lead to a faster propagation of shear waves and therefore a higher shear wave velocity (Ling, Li et al. 2016, Nenadic, Mynderse et al. 2016). This is due to formalin producing more cross-linking of tissue and increase the molecule lengths that would provide an easier path for wave propagation and thereby increase shearwave velocity and decrease attenuation (Werner, Chott et al. 2000). The lower starting velocity of 0.50 m/s as opposed to the value near 1.00 m/s in the original ex vivo tissue study was most likely due to the freezing and thawing of the samples as the formalin study was not conducted within a short period of time following the first ex vivo study. The trend in phase attenuation was shown in Fig. 5(c) and shows a gradual decrease in measured values. Pancreas one, two, and three showed phase attenuation percent differences and standard deviations between 0 and 100 minutes of 63.2 ± 21.0%, 64.8 ± 18.1%, and 55.8 ± 13.0%, respectively. As the tissue stiffens, the shear waves, once created, have less impediment for propagation and that is manifested as a lower shear wave attenuation. Therefore as the pancreas samples stiffened from the formalin the distance the excitation signal can propagate decreases.

Based on the results in the ex vivo study and formalin study, AMUSE was evaluated in healthy human subjects. The phase velocity and group velocity measurements were shown in Fig. 7, as well as the phase attenuation in Fig. 8 with the mean values and standard deviations reported in Table 2. The phase velocity measurements showed an increase of approximately 0.30–0.60 m/s for the head, body, and tail of the pancreas compared to the ex vivo results. The group velocity measurements showed an increase of approximately 0.80–1.00 m/s for the head, body, and tail of the pancreas when compared to the ex vivo results. This could be due to different environmental factors such as perfusion of the in vivo pancreas. The phase attenuation values showed no noticeable difference between the ex vivo and in vivo studies.

A few studies have examined the use of ultrasound- and MR-based elastography applied in the pancreas. A recent study has shown 100% sensitivity in diagnosing acute pancreatitis in 88 patients using acoustic radiation force impulse (ARFI)shear wave elasticity imaging (SWEI) (Göya, Hamidi et al. 2014). Their results showed a range of velocities in the head, body, and tail of healthy patients of 1.08–1.16 m/s, 1.14–1.20 m/s, and 1.08–1.12 m/s, respectively. The results of patients with chronic pancreatitis showed ranges for the head, body, and tail of 1.92–3.14 m/s, 2.01–3.43 m/s, and 1.77–2.74 m/s, respectively. Another study confirmed the feasibility of using ARFI induced SWEI measurements as a modality in the diagnosis of chronic pancreatitis. They reported shear wave velocities that were significantly higher in patients with chronic pancreatitis than in healthy patients. Their results showed shear wave velocities in the head, body, and tail of healthy volunteers of 1.23 ± 0.34 m/s, 1.30 ± 0.34 m/s, and 1.24 ± 0.50 m/s, respectively. They also reported velocities in the head, body, and tail of patients with chronic pancreatitis of 1.65 ± 0.71 m/s, 2.09 ± 1.03 m/s, and 1.68 ± 0.84 m/s, respectively (Yashima, Sasahira et al. 2012). Zaro, et al. studied 37 healthy subjects using SWEI to determine reference values for shear wave elastography in the pancreas. Their results reported mean shear wave velocity values of 1.22 m/s in the head, 1.23 m/s in the body, and 1.19 m/s in the tail (Zaro, Lupsor-Platon et al. 2016). Xie, et al. studied 44 subjects with acute pancreatitis against 210 healthy subjects using Virtual Touch Quantification (VTQ). Their results in the healthy group show shear wave velocities of 1.18 ± 0.23 m/s and 1.21 ± 0.20 m/s in the head and body, respectively. The acute pancreatitis group showed velocities of 1.18 ± 0.20 m/s and 1.25 ± 0.19 m/s in the head and body, respectively (Xie, Zou et al. 2015). These reported results show that the use of SWEI is effective in producing discernable shear wave velocity values between healthy patients and those with chronic pancreatitis. The study by Xie, et al. evaluated acute pancreatitis but showed that ARFI seemed to fail at distinguishing between healthy tissue and tissue with acute disease.

Magnetic resonance elastography (MRE) has been applied in the pancreas and the values of shear modulus are typically reported. For the purposes of comparing with the results in this study we converted the shear modulus values reported in the following studies to shear wave velocity assuming a mass density of 1000 kg/m3 (c=μ/ρ). Itoh et al. studied eight healthy volunteers as well as 25 subjects with focal pancreatic lesions to test elastic belt bracing of the abdomen with the goal of improving miscalculations in the pancreas. Their results showed that healthy individuals had mean shear wave velocity values of 1.54 m/s in the head, 1.57 m/s in the body, and 1.61 m/s in the tail of the pancreas. The mean shear wave velocity in pancreatic lesions were 2.46 m/s in the head, 2.36 m/s in the body, and 2.43 m/s in the tail of the pancreas. They also showed shear wave velocity values of 2.46 m/s in pancreatic cancers (Itoh, Takehara et al. 2016). An, et al. studied five healthy volunteers, five patients with chronic pancreatitis, and three patients with pancreatic ductal adenocarcinoma (PDAC) using 3D echo planar imaging MRE. Their results showed mean shear wave velocity values of 1.05 m/s for healthy tissue, 1.24 m/s for chronic pancreatitis, and 1.70 m/s for pancreatic ductal adenocarcinoma PDAC (An, Shi et al. 2016). Shi et al. studied 20 healthy volunteers using 3D echo planar imaging MRE to determine stiffness values in a healthy pancreas. Their results showed mean shear wave velocity values over the entire pancreas of 1.07 m/s at 40 Hz and 1.45 m/s at 60 Hz (Shi, Glaser et al. 2014).

When comparing the results of the ex vivo pancreas study to the results from healthy subjects in other papers, there is a clear similarity in values. Despite the different environmental factors between the ex vivo and in vivo results the comparable values confirmed the ability to evaluate this method with human volunteers. The studies using SWEI reported shear wave velocities with a range of 1.08–1.30 m/s for healthy tissue and studies using MRE reported slightly higher shear wave velocities with a range of 1.05–1.54 m/s. When scanning tissue with chronic pancreatitis, SWEI reported a range of shear wave velocities of 1.65–2.09 m/s. MRE reported a shear wave velocity of 1.24 m/s and a shear wave velocity range of 2.36–2.46 m/s for pancreatic lesions which are a possible result of chronic pancreatitis. Some of the MRE reports analyzed tissues diseased with cancer and reported a shear wave velocity range of 1.70–2.46 m/s. In comparison AMUSE gave a shear wave velocity range of 1.76–1.91 m/s in healthy subjects. This range is taken from the group velocities reported in Table 2. The range of values found in this study is higher than the range reported using SWEI or MRE by approximately 0.61–0.68 m/s and 0.52–0.67 m/s, respectively. This increase in the shear wave velocity numbers could be due to a factor in the processing of the shear wave data.

These comparisons need to be taken in light that the frequency ranges for the reported data may be different between this study and those with MRE and SWEI. The phase velocity and attenuation were evaluated at 100 Hz, which is close to the 60 Hz typically used for MRE. The SWEI results have frequency content that is unknown, but a group velocity is reported, so we find it relevant to compare to our group velocity and phase velocity results. The MRE and SWEI results provide a relevant range with MRE evaluations representing values at the lower end of the range and SWEI near 100 Hz or slightly above.

A difference is the use of group velocity as opposed to phase velocity as phase velocity only reports the shear wave velocity at one specific frequency whereas group velocity takes an average of the shear wave velocities throughout the entire range of frequencies. In the case of the ex vivo study, the phase velocity at 100 Hz was higher than the group velocity. Typically, if the material is viscoelastic, the phase velocity will increase with frequency. In this case, it would indicate that the phase velocities used to determine the group velocity were lower than 100 Hz, and this was confirmed through evaluation of the two-dimensional Fourier transform. In the ex vivo case, the peak of the frequency-domain representation was typically between 50–80 Hz. In the case of the in vivo human study, the phase velocity at 100 Hz was lower than the group velocity. This indicates that the phase velocities used to determine the group velocity are higher than 100 Hz. For the in vivo cases, there was more variability in the frequency range, but from the two-dimensional Fourier transform magnitude distributions, we did find that the magnitude did typically extend past 100–150 Hz. When considering these effects, the group velocities in healthy subjects would be lowered to within the range seen in ARFI and MRE, therefore confirming the effectiveness of AMUSE to produce comparable shear wave velocity results to the most commonly used methods for evaluating the elastography of the pancreas.

The AMUSE method provided new information related to the viscoelasticity of the pancreas including the phase velocity, and for the first time shear wave attenuation. The shear wave attenuation and phase velocity have been used in concert in the evaluation of acute cellular rejection in liver transplants (Nenadic, Qiang et al. 2017) and could be used in an advantageous manner for patients with pancreatitis as both conditions involve tissue inflammation.

There were a few limitations of this study. We only evaluated the shear wave phase velocity and attenuation at one frequency, 100 Hz. In future work, we would tailor the acoustic radiation force to try and obtain a more broadband shear wave response. Additionally, we only studied 12 human subjects. We did obtain consistent results among the different subjects, but a larger cohort and cohorts with pancreatitis or pancreatic cancer need to be studied to fully evaluate the efficacy of using AMUSE for assessment of the pancreas.

Conclusion

AMUSE is a novel ultrasound method that uses shear wave velocity as well as attenuation to characterize both elasticity and viscosity of the tissue. AMUSE was tested in ex vivo porcine pancreas samples that were also submerged in formalin to simulate stiffening. There was a noticeable difference between normal and stiffened tissue shear wave velocity values. This method was also used in healthy human subjects. To further test this method, scanning subjects with confirmed pancreatitis and pancreatic cancer is required. This will aid in evaluation to determine the usefulness of phase velocity and/or shear wave attenuation for the differentiation of patients with pancreatic pathologies.

References

  1. Amador C, et al. (2017). “Improved shear wave group velocity estimation method based on spatiotemporal peak and thresholding motion search.” IEEE Transactions on Ultrasonics. Ferroelectrics, and Frequency Control 64(4): 660–668. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Amador C, et al. (2016). “Ultrasonic method to characterize shear wave propagation in micellar fluids.” The Journal of the Acoustical Society of America 140(3): 1719–1726. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. An H, et al. (2016). “Test-retest reliability of 3D EPI MR elastography of the pancreas.”Clinical Radiology 71(10): 1068.e1067–1068.e1012. [DOI] [PubMed] [Google Scholar]
  4. Bercoff J, et al. (2004). “Supersonic shear imaging: a new technique for soft tissue elasticity mapping.” IEEE Trans Ultrason Ferroelectr Freq Control 51(4): 396–409. [DOI] [PubMed] [Google Scholar]
  5. Bernal M, et al. (2011). “Material property estimation for tubes and arteries using ultrasound radiation force and analysis of propagating modes.” Journal of the Acoustical Society of America 129(3): 1344–1354. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Budelli E, et al. (2017). “A diffraction correction for storage and loss moduli imaging using radiation force based elastography.” Physics in Medicine and Biology 62(1): 91–106. [DOI] [PubMed] [Google Scholar]
  7. Chen S, et al. (2009). “Shearwave dispersion ultrasound vibrometry (SDUV) for measuring tissue elasticity and viscosity.” IEEE Trans Ultrason Ferroelectr Freq Control 56(1): 55–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Chen S, et al. (2009). “Shearwave dispersion ultrasound vibrometry (SDUV) for measuring tissue elasticity and viscosity.” IEEE Transactions on Ultrasonics Ferroelectrics and Frequency Control 56(1): 5562. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chu L, et al. (2017). “Diagnosis and Detection of Pancreatic Cancer.” Cancer Journal 23(6): 333–342. [DOI] [PubMed] [Google Scholar]
  10. Deffieux T, et al. (2009). “Shear wave spectroscopy for in vivo quantification of human soft tissues visco-elasticity.” IEEE Transactions on Medical Imaging 28(3): 313–322. [DOI] [PubMed] [Google Scholar]
  11. Göya C, et al. (2014). “Use of Acoustic Radiation Force Impulse Elastography to Diagnose Acute Pancreatitis at Hospital Admission: Comparison With Sonography and Computed Tomography.” Journal of Ultrasound in Medicine 33(8): 1453–1460. [DOI] [PubMed] [Google Scholar]
  12. Herman BA and Harris GR (2002). “Models and regulatory considerations for transient temperature rise during diagnostic ultrasound pulses.” Ultrasound Med Biol 28(9): 1217–1224. [DOI] [PubMed] [Google Scholar]
  13. Hu JJ, et al. (2018). “Ultrasound Shear Wave Elastography as a Measure of Porcine Hepatic Disease in Right Heart Dysfunction: A Pilot Study.” Ultrasound in Medicine & Biology 44(11): 2393–2399. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Itoh Y, et al. (2016). “Feasibility of magnetic resonance elastography for the pancreas at 3T.” Journal of Magnetic Resonance Imaging 43(2): 384–390. [DOI] [PubMed] [Google Scholar]
  15. Jiang Y, et al. (2015). “Characterization of the nonlinear elastic properties of soft tissues using the supersonic shear imaging (SSI) technique: Inverse method, ex vivo and in vivo experiments.” Medical Image Analysis 20(1): 97–111. [DOI] [PubMed] [Google Scholar]
  16. Johnson C, et al. (2014). “Acute Pancreatitis.” BMJ 349. [DOI] [PubMed] [Google Scholar]
  17. Kasai C, et al. (1985). “Real-time two-dimensional blood flow imaging using an autocorrelation technique.” IEEE Transactions on Sonics and Ultrasonics SU-32(3): 458–464. [Google Scholar]
  18. Kumar V, et al. (2018). “Viscoelastic parameters as discriminators of breast masses: Initial human study results.” PLOS ONE 13(10): e0205717. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Ling Y, et al. (2016). “Effects of fixation and preservation on tissue elastic properties measured by quantitative optical coherence elastography (OCE).” Journal of Biomechanics 49(7): 1009–1015. [DOI] [PubMed] [Google Scholar]
  20. Montaldo G, et al. (2009). “Coherent plane-wave compounding for very high frame rate ultrasonography and transient elastography.” IEEE Transactions on Ultrasonics Ferroelectrics and Frequency Control 56(3): 489–506. [DOI] [PubMed] [Google Scholar]
  21. Nenadic I, et al. (2016). “Noninvasive evaluation of bladder wall mechanical properties as a function of filling volume: potential application in bladder compliance assessment.” PLoS ONE 11(6): e0157818. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Nenadic IZ, et al. (2017). “Attenuation measuring ultrasound shearwave elastography and in vivo application in post-transplant liver patients.” Physics in Medicine and Biology 62(2): 484–500. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Palmeri ML, et al. (2008). “Quantifying hepatic shear modulus in vivo using acoustic radiation force.” Ultrasound in Medicine and Biology 34(4): 546–558. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Pelosi E, et al. (2017). “Pancreatic Cancer: Molecular Characterization, Clonal Evolution and Cancer Stem Cells.” Biomedicines 5(65): 55. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Rouze NC, et al. (2017). “Accounting for the spatial observation window in the 2-D Fourier transform analysis of shear wave attenuation.” Ultrasound in Medicine & Biology 43(10): 2500–2506. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Rouze NC, et al. (2015). “An analytic, Fourier domain description of shear wave propagation in a viscoelastic medium using asymmetric Gaussian sources.” The Journal of the Acoustical Society of America 138(2): 1012–1022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Sarvazyan A, et al. (2011). “An Overview of Elastography - an Emerging Branch of Medical Imaging.” Current medical imaging reviews 7(4): 255–282. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Shi Y, et al. (2014). “Feasibility of using 3D MR elastography to determine pancreatic stiffness in healthy volunteers ” Journal of Magnetic Resonance Imaging 41(2): 369–375. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Siegel RL, et al. (2017). “Cancer statistics, 2017.” CA: A Cancer Journal for Clinicians 67(1): 7–30. [DOI] [PubMed] [Google Scholar]
  30. Silva-Vaz P, et al. (2019). “Murine Models of Acute Pancreatitis: A Critical Appraisal of Clinical Relevance.” IJMS 20(2794): 19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Stram M, et al. (2016). “Chronic Pancreatitis.” Surgical Pathology Clinics 9(4): 643–659. [DOI] [PubMed] [Google Scholar]
  32. Tanter M, et al. (2008). “Quantitative assessment of breast lesion viscoelasticity: Initial clinical results using supersonic shear imaging.” Ultrasound in Medicine and Biology 34(9): 1373–1386. [DOI] [PubMed] [Google Scholar]
  33. Tanter M and Fink M (2014). “Ultrafast imaging in biomedical ultrasound.” IEEE Trans Ultrason Ferroelectr Freq Control 61(1): 102–119. [DOI] [PubMed] [Google Scholar]
  34. Walling A and Freelove R (2017). “Pancreatits and Pancreatic Cancer.” Primary Care: Clinics in Office Practice 44(4): 609–620. [DOI] [PubMed] [Google Scholar]
  35. Werner M, et al. (2000). “Effect of Formalin Tissue Fixation and Processing on Immunohistochemistry.” The American Journal of Surgical Pathology 24(7): 1016–1019. [DOI] [PubMed] [Google Scholar]
  36. Xie J, et al. (2015). “A Preliminary Investigation of Normal Pancreas and Acute Pancreatits Elasticity Using Virtual Touch Tissue Quantification (VTQ) Imaging.” Medical Science Monitor 21: 1693–1699. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Yashima Y, et al. (2012). “Acoustic radiation force impulse elastography for noninvasive assessment of chronic pancreatitis.” Journal of Gastroenterology 47(4): 427–432. [DOI] [PubMed] [Google Scholar]
  38. Zaro R, et al. (2016). “The pursuit of normal reference values of pancreas stiffness by using Acoustic Radiation Force Implues (ARFI) elastography.” Medical Ultrasongraphy 18(4): 425–430. [DOI] [PubMed] [Google Scholar]

RESOURCES