ROI Shear-wave Elastography Measurement Method

We read with great interest the article by Arda et al. [1], “Quantitative Assessment of Normal Soft-Tissue Elasticity Using Shear-Wave Ultrasound Elastography,” in the September 2011 issue of the AJR. The authors quantitatively measured the elasticity (in kilopascals) of various normal tissues in healthy volunteers to create a formal basis of reference. Promising ultrasound techniques include the following: Elastography tracks tissue motion during compression to obtain strain estimates, sonoelastography uses color Doppler signal to image tissue motion in response to probe vibrations, and tracking of shear wave propagation through tissue provides elastic modulus values. A major drawback of color qualitative elastography is its high intra- and interobserver variability associated with iterative ultrasound probe compression [2].

Compared with other elasticity imaging techniques, shear-wave elastography can measure local mechanical tissue properties almost independently from adjacent tissues. Additionally, shear-wave elastography does not require external deformation means or a vibration source. Shear-wave elastography gives real-time quantitative information about tissue stiffness by measuring and displaying local tissue elasticity on a color-coded map (Fig. 1). Thus, shear-wave elastography is theoretically insensitive to target size and allows interobserver reproducibility, quality, and accuracy using automatic tissue stiffness measurement in regions of interest (ROIs) [1, 3, 4]. However, recent studies by Arda et al. [1] and Sebag et al. [4] lack precise technical criteria to assess healthy tissue elasticity [1, 4]—namely, the number of ultrasound elastographic acquisitions performed per tissue analysis; matching of ROI values assessed by two independent observers; and the ROI shear-wave elastography measurement method, including location, size, number, and acquisition plane. This last issue is particularly important in anisotropic tissues, such as the thyroid gland or even more in the muscle tendons. In the study by Arda et al., variations of elasticity values approached 100% between the axial and longitudinal planes for the Achilles tendon, but the plane that was chosen as a formal reference was not specified.

Fig. 1

—Single-shot 3D shear-wave elastography in 50-year-old patient with healthy thyroid.

A–D, Axial (A), longitudinal (B), and coronal (C) plane ultrasound images are secondarily reformatted using single source of data. Three-dimension acquisition of right thyroid lobe is shown (D). The region of interest (ROI) shear-wave elastography calculation is made secondarily on each plane by placing ROI in target area on color maps of elastography images. Elastography values (mean stiffness, maximum, minimum, and SD) are displayed instantaneously on each reformatted plane.

Furthermore, variations in the ultrasound probe pressure also might induce significant variations in the ROI shear-wave elastography calculation. The probe pressure helps to chase the bowel air to target the pancreas but may induce organ elasticity assessment errors as it can in superficial organs (Fig. 2). Determining the ROI ratio of the target tissue to the reference adjacent tissue values (muscle, fat, normal adjacent tissue versus pathology area) should thus improve the reproducibility of shear-wave elastography.

Using a 3D single-shot volume acquisition of the target organ with secondary calculation limits the risk of error in ROI calculation of probe pressure variability compared with successive axial and longitudinal plane calculations (Fig. 1). Moreover, establishing a ratio is a way to compensate for ROI calculation variability along the axis related to probe pressure.

In conclusion, to limit intra- and interobserver variability of quantitative elastography [3, 4], we emphasize the necessity of establishing a wide range of normal and pathologic tissue shear-wave elastography values measured along the same (longitudinal or axial) axis, performed—if possible—on a single 3D acquisition and using a systematic ROI ratio. We advocate an accurate definition of the formal assessment technique of ROI shear-wave elastography calculation.

Fig. 2

—Variations of region of interest (ROI) calculation as function of probe pressure in 50-year-old patient with healthy thyroid (same patient as inFig. 1).

A and B, Axial (A) and longitudinal (B) axis images with same patient and same ROI location and size as in Figure 1 show ROI values vary from one to double (no compression, 1; mild compression, 2; strong probe compression, 3). Q-Box quantification tool manufactured by Medivision Benelux.

Next Section

• © American Roentgen Ray Society

Previous Section