Elastography imaging of small animal oncology models: a feasibility study

Application of Elastography for the Noninvasive Assessment of Biomechanics in Engineered Biomaterials and Tissues

The elastic properties of engineered biomaterials and tissues impact their post-implantation repair potential and structural integrity, and are critical to help regulate cell fate and gene expression. The measurement of properties (e.g., stiffness or shear modulus) can be attained using elastography, which exploits noninvasive imaging modalities to provide functional information of a material indicative of the regeneration state. In this review, we outline the current leading elastography methodologies available to characterize the properties of biomaterials and tissues suitable for repair and mechanobiology research. We describe methods utilizing magnetic resonance, ultrasound, and optical coherent elastography, highlighting their potential for longitudinal monitoring of implanted materials in vivo, in addition to spatiotemporal limits of each method for probing changes in cell-laden constructs. Micro-elastography methods now allow acquisitions at length scales approaching 5-100 μm in two and three dimensions. Many of the methods introduced in this review are therefore capable of longitudinal monitoring in biomaterials and tissues approaching the cellular scale. However, critical factors such as anisotropy, heterogeneity and viscoelasity-inherent in many soft tissues-are often not fully described and therefore require further advancements and future developments.

Transient micro-elastography: A novel non-invasive approach to measure liver stiffness in mice

AIM: To develop and validate a transient micro-elastography device to measure liver stiffness (LS) in mice.

METHODS: A novel transient micro-elastography (TME) device, dedicated to LS measurements in mice with a range of measurement from 1-170 kPa, was developed using an optimized vibration frequency of 300 Hz and a 2 mm piston. The novel probe was validated in a classical fibrosis model (CCl₄) and in a transgenic murine model of systemic amyloidosis.

RESULTS: TME could be successfully performed in control mice below the xiphoid cartilage, with a mean LS of 4.4 ± 1.3 kPa, a mean success rate of 88%, and an excellent intra-observer agreement (0.98). Treatment with CCl₄ over seven weeks drastically increased LS as compared to controls (18.2 ± 3.7 kPa vs 3.6 ± 1.2 kPa). Moreover, fibrosis stage was highly correlated with LS (Spearman coefficient = 0.88, P < 0.01). In the amyloidosis model, much higher LS values were obtained, reaching maximum values of > 150 kPa. LS significantly correlated with the amyloidosis index (0.93, P < 0.0001) and the plasma concentration of mutant hapoA-II (0.62, P < 0.005).

CONCLUSION: Here, we have established the first non-invasive approach to measure LS in mice, and have successfully validated it in two murine models of high LS.

3D strain imaging using a rectilinear 2D array

Under mechanical compression, tissue movements are inherently three-dimensional. 2-D strain imaging can suffer from decorrelation noise caused by out-of-plane tissue movement in elevation. With 3-D strain imaging, all tissue movements can be estimated and compensated, hence minimizing out-of-plane decorrelation noise. Promising 3-D strain imaging results have been shown using 1-D arrays with mechanical translation in elevation. However, the relatively large slice thickness and mechanical translation can degrade image quality. Using 2-D arrays, an improved elevational resolution can be achieved with electronic focusing. Furthermore, scanning with 2-D arrays is also done electronically, which eliminates the need for mechanical translation. In this paper, we demonstrate the feasibility of 3-D strain imaging using a 4 cm x 4 cm ultrasonic sparse rectilinear 2-D array operating at 5MHz. The signal processing combinations of 2-D or 3-D beamforming followed by 2-D or 3-D strain imaging are studied and compared to each other to evaluate the performance of our 3-D strain imaging system. 3-D beamforming followed by 3-D strain imaging showed best performance in all experiments.

3D prostate elastography: algorithm, simulations and experiments

A multi-resolution hybrid strain estimator is presented. The estimator is locally initialized by the B-mode tracking stage. Nonlinear and linear stretching regimes are applied in successive RF tracking stages for refining the estimated axial and lateral displacements. A staggering operator is used to derive the strain images from the reconstructed axial displacements. Simulations and experiments, conducted at a center frequency of 12 MHz, 40% fractional bandwidth, on a 128 element transducer with 0.2 mm pitch, with elastographic window length of 2 mm and overlap of 90%, demonstrate a 3-6 dB improvement in the elastographic contrast-to-noise ratio over the results obtained using conventional multi-stage stretching based strain estimators. The average image cross-correlation coefficient obtained using the proposed algorithm was improved by 6-8%. 3D elastographic simulations conducted to study the performance of a 3D elastographic imaging framework predict achievable axial and lateral resolutions of approximately five and ten wavelengths, respectively. A close correspondence between inclusions reconstructed from experimental elastograms and the known physical shape of actual 3D inclusions demonstrates the potential application of 3D elastography for identifying and classifying the detected lesions (invisible in sonograms) on the basis of their shape.

Tissue Doppler and strain imaging for evaluating tissue elasticity of breast lesions

RATIONALE AND OBJECTIVES

Sonoelastography depicts the intrinsic elastic properties of a tissue which are characterized by the strain applied to achieve tissue deformation and the velocity at which tissue deformation occurs. The present study served to investigate whether the specificity of B-mode ultrasound (US) can be improved by combining B-mode imaging with tissue Doppler imaging (TDI) and offline analysis of tissue strain imaging (TSI).

MATERIALS AND METHODS

Fifty women, 25 with malignant and 25 with benign focal breast lesions, were examined by US with a linear transducer (9 MHz, Aplio, Toshiba, Otawara, Japan). B-mode US views of the lesions were overlaid with color-coded TDI information and area quotients (AQ = area B-mode view/area TDI) were calculated. TSI views were reconstructed offline from the source data. This was done by placing a region of interest (ROI) in the target lesion and color-encoded display of the information. In addition, tissue elasticity was evaluated using a scale of 1-5 corresponding to the BI-RADS categories. Maximum strain (strain factor, SF) was determined in the ROI. All patients also underwent mammography. Sensitivities and specificities were calculated and statistical analysis was performed using Wilcoxon's test.

RESULTS

Sensitivity/specificity was 96%/68% for B-mode US, 100%/40% for combined B-mode US and mammography, and 96%/80% for TSI. The AQ of benign and malignant lesions was significantly different (p = .00008) as was the difference in SF (p = .0004). The readers considered TSI a feasible technique.

CONCLUSION

Evaluation of elasticity based on the quantification of strain factors improves characterization of focal breast lesions, especially the differentiation of BI-RADS 3 and 4 lesions. Surprisingly, significant results in characterizing breast lesions were obtained with the simple technique of TDI, showing a lower tissue displacement in malignant cases.

Real-time sonoelastography of the cervix: tissue elasticity of the normal and abnormal cervix

RATIONALE AND OBJECTIVES

First study to investigate the basic tissue elastic properties of the cervix in pre- and postmenopausal healthy women and to compare these normal findings with the results in a group of patients with focal pathology of the cervix.

MATERIALS AND METHODS

A total of 113 patients underwent transvaginal ultrasound, among them 24 with cervical pathology. The real-time elastography (Hitachi) information was color-coded and superimposed on the B-mode scan. The elastography images were analyzed by means of a software tool to identify thresholds for the colors red (soft), blue (hard), and green (medium hard), and the percentages of the three colors of the total area were determined. The results were correlated with age. In addition, scans were evaluated subjectively on an analogue scale from 1 (definitely normal) to 5 (definitely abnormal). Statistical analysis was performed using Anova, Wilcoxon's test, and Pearson's correlation.

RESULTS

Computer-assisted generation of the color spectrum showed green to be predominant in both the normal group (67+/-13 %) and in the group with cervical pathology (64+/-15 %) without a significant difference between both groups (p=>0.05). Significant differences (p<0.05) in the blue color spectrum (hard tissue) were found between the 13 cervical tumor patients (34+/-15 %) and the normal group (26+/-13 %) but not between the CIN patients and normal women (19+/-12 %) (p>0.05). Subjective tumor characterization also showed significant differences (p<0.05) among the groups and good correlation with the histologic diagnosis (r2=0.744). There were no significant changes in color distribution with patient age (p>0.05).

CONCLUSION

Computer-assisted and subjective evaluation of cervical elastography allows differentiation of malignancy from normal findings. CIN cannot be identified with this modality. Elastographically, cervical tissue is of medium hardness and does not change with age.

Real-time sonoelastography performed in addition to B-mode ultrasound and mammography: improved differentiation of breast lesions?

RATIONALE AND OBJECTIVES

The goal of the present study was to compare the sensitivity and specificity of elastography with that of B-mode ultrasound (US) and mammography.

MATERIALS AND METHODS

A total of 300 patients with histologically confirmed breast lesions (168 benign, 132 malignant) were included. Evaluation was by means of the three-dimensional finite-element method. The data are color-coded and superimposed on the B-mode US scan. The images were evaluated by two independent readers. The results were compared with mammography, histology, and the data obtained by previous US investigations. Sensitivities and specificities were calculated.

RESULTS

Sensitivity and specificity in the differentiation of benign and malignant lesions were 87% and 85%, respectively, for mammography and 94% and 83% for B-mode US. The two examiners were in very good agreement in their evaluation of the elastograms (kappa: 0.86). Elastography had a sensitivity of 82% and a specificity of 87%. Elastography was superior to B-mode US in diagnosing Breast Imaging Reporting and Data System (BI-RADS) 3 lesions (92% vs. 82% specificity) and in lipomatous involution (80% vs. 69% specificity).

CONCLUSION

Elastography in breast lesions showed a higher specificity and a lower sensitivity in comparison with B-mode sonography. Elastography may be beneficial in BI-RADS 3 lesions and in lipomatous involution.

Optical coherence elastography of engineered and developing tissue

Biomechanical elastic properties are among the many variables used to characterize in vivo and in vitro tissues. Since these properties depend largely on the micro- and macroscopic structural organization tissue, it is crucial to understand the mechanical properties and the alterations that occur tissues respond to external forces or to disease processes. Using a novel technique called coherence elastography (OCE), we mapped the spatially distributed mechanical displacements strains in a representative model of a developing, engineered tissue as cells began to proliferate attach within a three-dimensional collagen matrix. OCE was also performed in the complex tissue of the Xenopus laevis (African frog) tadpole. Displacements were quantified a cross-correlation algorithm on pre- and postcompression images, which were acquired using coherence tomography (OCT). The images of the engineered tissue were acquired over a 10-development period to observe the relative strain differences in various regions. OCE was able differentiate changes in strain over time, which corresponded with cell proliferation and matrix as confirmed with histological observations. By anatomically mapping the regional variation stiffness with micron resolution, it may be possible to provide new insight into the complex by which engineered and natural tissues develop complex structures.

In vivo elastographic investigation of ethanol-induced hepatic lesions

Ethanol-induced hepatic lesions were investigated in swine for in vivo use as a strain imaging animal model. Lesions (n = 25) were induced by injecting ethanol (doses 0.33 to 2.0 mL) directly into the surgically exposed liver at depths of 12, 15 or 25 mm. Lesions were imaged with a modified HDI 1000 scanner (Philips Medical Systems, Bothell, WA, USA). The elastograms (n = 91) characterized lesions as being areas harder than the surrounding soft hepatic tissue. Elastographic lesion sizes and the corresponding injected ethanol dose used to induce the lesions were shown to be statistically significant (r(2) = 0.22; p = 0.029) using a linear regression analysis. Additionally, lesion depth was shown to be statistically insignificant (r(2) < 0.12; p > 0.10) when regressed against elastographic lesion size. An analysis of elastographic and gross pathology lesion sizes indicated no correlation (r(2) < 0.01; p = 0.973). Subsequently, lesion types were sorted by size and regression lines were computed from quasilinear regions of the corresponding run charts. Trend lines indicate a four-to-three size relationship between the selected elastographic and pathology lesion sizes. Comparison of elastogram lesion sizes from two independent observers using a paired t-test resulted in no statistically significant difference (p = 0.14). In conclusion, ethanol-induced hepatic lesions in swine is a suitable animal model for evaluation of strain-based imaging systems, due to the ease of generation and repeatability.