- Email: [email protected]
Recent technological advancements in breast ultrasound
Ultrasonics 70 (2016) 183–190
Contents lists available at ScienceDirect
Ultrasonics journal homepage: www.elsevier.com/locate/ultras
Recent technological advancements in breast ultrasound John R. Eisenbrey ⇑, Jaydev K. Dave, Flemming Forsberg Thomas Jefferson University, Department of Radiology, Division of Ultrasound, 132 South 10th St., Philadelphia, PA 19107, United States
a r t i c l e
i n f o
Article history: Received 16 March 2016 Received in revised form 20 April 2016 Accepted 24 April 2016 Available online 25 April 2016 Keywords: Breast ultrasound Automated breast ultrasound Elastography Contrast-enhanced ultrasound
a b s t r a c t Ultrasound is becoming increasingly common as an imaging tool for the detection and characterization of breast tumors. This paper provides an overview of recent technological advancements, especially those that may have an impact in clinical applications in the field of breast ultrasound in the near future. These advancements include close to 100% fractional bandwidth high frequency (5–18 MHz) 2D and 3D arrays, automated breast imaging systems to minimize the operator dependence and advanced processing techniques, such as those used for detection of microcalcifications. In addition, elastography and contrast-enhanced ultrasound examinations that are expected to further enhance the clinical importance of ultrasound based breast tumor screening are briefly reviewed. These techniques have shown initial promise in clinical trials and may translate to more comprehensive clinical adoption in the future. Ó 2016 Elsevier B.V. All rights reserved.
Contents 1. 2. 3. 4. 5. 6. 7.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 2D and 3D transducer technologies . . . . . . . . Whole breast ultrasound screening . . . . . . . . Image acquisition and processing techniques Elastography . . . . . . . . . . . . . . . . . . . . . . . . . . . Contrast-enhanced ultrasound . . . . . . . . . . . . Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
1. Introduction Breast cancer is the single most commonly occurring cancer in women in the United States with over 249,000 cases, and second most common cause of cancer death in women (41,152 deaths expected in 2016) [1]. Breast cancer screening is an important step for the early diagnosis of malignancy as 5 year survival rates vary dramatically by stage of initial diagnosis (as high as 100% at stage I (see [2] for staging criteria) and roughly 25% when diagnosed at stage IV) [1]. X-ray mammography is currently the primary imaging screening tool and recommended in the United States for women 40 years and older [3], although considerable controversy currently surrounds the appropriate age to begin screening [4]. Despite its advantages mammography suffers from poor overall ⇑ Corresponding author. E-mail address: [email protected] (J.R. Eisenbrey). http://dx.doi.org/10.1016/j.ultras.2016.04.021 0041-624X/Ó 2016 Elsevier B.V. All rights reserved.
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
183 184 184 184 185 187 189 189
specificity, resulting in false-positive rates of 65–90% [5,6]. Additionally, mammography is often not adequate in the 20–50% of patients with dense breasts [5]. Ultrasound may overcome both of these limitations. Breast density does not inhibit ultrasound waves in the breast making breast ultrasound imaging a useful technique in these women [7]. Breast ultrasound has also been shown to be a useful adjunct for the characterization of masses post-mammography using an ultrasound-based Breast Imaging Reporting and Data Systems (BIRAD) scoring system. Breast ultrasound post mammography has shown to result in a down staging of roughly 20–40% of BIRADS 4A cases and of roughly 15–25% of BIRADS 3 cases, thus limiting the number of patients requiring biopsy [7]. Unnecessary biopsies are undesirable due to patient discomfort and anxiety, risk of infection, and cost of the procedure. When biopsy is still required, ultrasound is also routinely used for image guided breast mass biopsies [7].
184
J.R. Eisenbrey et al. / Ultrasonics 70 (2016) 183–190
While ultrasound is already playing an important clinical role in breast imaging, overall specificities are still relatively low (around 66% based on a retrospective review of 761 cases using only Bmode imaging [8]). Hence, numerous techniques are currently in development to improve ultrasound’s prominence in this application. This paper will serve as an overview of emerging concepts in the field, including high frequency imaging and processing approaches, automated volumetric breast ultrasound systems, breast elastography, and contrast-enhanced breast ultrasound. 2. 2D and 3D transducer technologies High frequency, broad bandwidth transducers can produce excellent detail resolution at the relatively shallow depths (typically < 4 cm) encountered in breast imaging. Currently available commercial systems use linear arrays operating around 10– 14 MHz with close to 100% bandwidth ranging from 5 to 18 MHz. With a typical (lateral) resolution close to 2k, a 10 MHz transmit would yield resolution of approximately 300 lm. Thus, such arrays are capable of providing superior resolution at multiple depths by selecting the best possible compromise between penetration depth, which is dependent on attenuation of the pressure amplitude probing the tissue, and the wavelength dependent resolution. Examples of bandwidths in excess of 150% with center frequencies of 9–15 MHz have been reported using Polyvinylidene Fluoride (PVDF) as well as capacitive microfabricated ultrasound transducer (cMUT) probes [9–11]. A major goal in transducer development has been the construction of electronic volumetric arrays, which would allow for simultaneous beam-steering in the axial, lateral and azimuthal directions and the acquisition of full volumetric data sets (focused in all three dimensions) [10–13]. Moreover, volumetric arrays would permit true vector velocity flow imaging, since a volume data set allows flow to be interrogated at multiple angles, which could be combined to determine the true 3D flow vector. The challenge in construction of volumetric arrays has always been the large number of elements (10,000+) and, while multiplexers in the transducer handle have limited the number of wires necessary for use it will not be trivial to construct volumetric arrays operating at the higher frequencies required in breast imaging. Nonetheless, commercial volumetric phased as well as linear arrays for use in echocardiography are now available from several manufacturers [12,13]. These matrix arrays can produce real time 3D imaging (over 20 volumes per second), but they operate at frequencies too low to be suitable for breast imaging (i.e., <7 MHz; resolution of approximately 500 lm). 3. Whole breast ultrasound screening Whole breast ultrasound screening studies to date have focused on women with elevated breast cancer risk and/or dense breast tissue [14,15]. When comparing mammography alone to screening with ultrasound and mammography the largest study to date (the ACRIN 6666 trial) involving over 2800 women found an additional 1.1 to 7.2 cancers per 1000 high-risk women; albeit with a marked increase in the number of false positives [14]. This group of investigators also compared ultrasound screening alone to conventional mammographic screening and concluded that cancer detection rates were comparable [15]. As before, false positives were more common with whole breast ultrasound screening. However, screening examinations can be time consuming and hand held transducers are by definition limited to a 2D field of view. These exams may take up to 30 min compared to roughly 10 min for a mammogram, although with less patient discomfort and no associated radiation. As an alternative, large field of view
systems capable of scanning an entire breast using very large linear arrays (14–15 cm in length) have been developed for screening purposes [16–18]. Automated 3D whole breast ultrasound systems allow the technologist to acquire high quality images (resolutions < 300 lm) in the standard mammographic views across three planes and have been shown to achieve the same diagnostic accuracy (>87%) as hand held ultrasound [16,17]. The FDA approved automated breast ultrasound in 2012 as an adjunct screening modality to mammography in asymptomatic women with dense breast tissue (breast density scores of 3–4, based on a mammography-indicated density of over 51%) for whom screening mammography findings are normal or benign. With depiction of the coronal plane, mass margins, shape, spiculations, and distortion associated with tissue retraction are visible [16–18].
4. Image acquisition and processing techniques Two image acquisition techniques – compound imaging and tissue harmonic imaging (THI) – have become the bread and butter processing techniques of clinical breast imaging [19–23]. In spatial compound imaging several (typically 3–9) ultrasound images are acquired at different angles of insonation and averaged to produce a single image. Averaging reduces speckle and improves the delineation of lateral borders, which results in increased conspicuity of low-contrast lesions, enhanced delineation of capsular margins and ducts, and better overall image quality when using spatial compounding compared to conventional imaging [19,20,23]. Similarly, frequency compounding uses multiple filters to create images from different received frequency bands. These images are subsequently averaged to suppress speckle, increase contrast resolution and improve penetration [21]. Frequency compounding is also less susceptible to the depth-related issues that affect spatial compounding. In THI mode, the scanner is configured to receive echoes at double the transmit (or fundamental) frequency. By optimizing the receive filters for the weak, nonlinear tissue signal components that arise once sound waves propagate some frequency dependent distance into the tissue, it is possible to reduce echoes from clutter and side lobes at the fundamental frequency, while improving contrast resolution and border delineation [22,23]. The grayscale contrast between glandular or fatty tissue and breast lesions is improved with THI compared with conventional breast ultrasound [22]. The simultaneous use of spatial compounding and THI increases overall image quality and lesion conspicuity, and is now recommended in routine clinical practice [23]. MicroPure (Toshiba America Medical Systems, Tustin, CA) is a new commercial ultrasound image processing technique that post-processes grayscale ultrasound images in order to improve the visualization of breast microcalcifications [24–26]. This software uses a filter technique, where each pixel is compared to the average brightness of the surrounding area, to detect locations where there is a characteristic change from surrounding area [24]. As the focus of MicroPure imaging is to detect microcalcifications in breast tissue, the filter kernel is optimized in the horizontal direction to detect only isolated points with higher brightness compared to the surrounding breast tissue [24–26]. An example of this technique being used to improve ultrasound visualization of microcalcifications in the breast is provided in Fig. 1. Independent investigations have established that MicroPure can detect more microcalcifications than standard ultrasound in women presenting with microcalcifications seen on mammography [24,25]. However, this method is not appropriate for a screening population as it does not possess the same level of sensitivity as mammography, and should be used in more focused applications such as for guiding a biopsy [26].
J.R. Eisenbrey et al. / Ultrasonics 70 (2016) 183–190
185
Fig. 1. Example imaging of MicroPure imaging being used to visualize breast microcalcifications. Dual imaging is used in which grayscale B-mode ultrasound is displayed in left side of the image, and MicroPure is displayed in blue on the right. Calcifications are identified by the brighter white echoes overlain on the blue image. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
5. Elastography Elastography based techniques convey information about stiffness of a lesion under consideration, with stiffer tissue generally being indicative of malignancy. Since early published literature on the use of ultrasound for elastography [27–29] several singleand multi- center studies have been conducted to evaluate the usefulness of elastography to characterize breast lesions, and the most recent guidelines for clinical use of elastography were published in 2015 by World Federation of Ultrasound in Medicine and Biology (WFUMB) and recommend elastography for the characterization of solid breast masses [30,31]. The overall principle of elastography depends on introducing mechanical excitation within a region of interest and measuring the induced disturbance. The induced disturbance is measured either as displacement within the field of view and this is referred to as strain elastography (SE) or as the velocity of an induced shear wave and this is referred to as shear wave elastography (SWE) [32,33,30,31]. Both SE and SWE are used for characterizing breast lesions. From the technological perspective for SE of breast lesions, the induced disturbance is compression either achieved manually or using a subject’s respiratory or cardiac pulsations, or via Acoustic Radiation Force Imaging (ARFI) wherein the ultrasound transducer is used to initiate a pulse resulting in disturbance. For SWE of breast lesions, again ARFI is used but the speed of the induced shear wave in the lesion is measured (SWE with manual compression is not performed). The measured shear wave speed (in m/s) can also be expressed as stiffness (in kPa) after conversion (assuming a Young’s modulus of 0.5) as shown in Fig. 2. The characterization of breast lesions as benign or malignant based on SE and SWE is based on the principle that benign lesions are relatively easily deformable and therefore, characterized by low stiffness values as compared to malignant lesions that exhibit increased stiffness values. The elastography images are interpreted based on qualitative, semi-quantitative and quantitative assessments [34–37]. In brief, these assessments involve characterization of lesions based on a lesion-stiffness dependent color coded images derived from SE or SWE and measurements based on grayscale and corresponding elastography images including ratios of lesion size and strain measurements in fat and lesion. Based on these assessments it has been shown that a major advantage of utilizing elastography for breast lesion characterization is the improvement in
specificity up to 20% (relative to B-mode ultrasound) to rule out biopsy specifically to distinguish lesions categorized as BI-RADS category 3 or 4A [35,37–40]. Two recent studies have shown that the performance of SE and SWE for differentiation between benign and malignant breast lesions is similar with equivalent areas under their receiver operating characteristic curves (AUC > 0.9) [41,42]. Another multi-center study involving subjects compared two SE techniques, namely, a qualitative scoring system for elastography images and assessment of strain ratios, and concluded that in their study population the diagnostic performance characterized by these two SE techniques was not different (area under the receiver operator characteristic (ROC) curve for both techniques was 0.86). [43]. Such data suggest that following guidelines as recommended by WFUMB [31], both SE and SWE techniques may yield similar performance. Preliminary results also indicate that breast lesion stiffness may be closely associated with lymph node involvement and therefore SE or SWE may also be useful in characterizing the metastatic nature of a malignant lesion [44]. Also, the microenvironment of benign and malignant cells is different and invasiveness associated with malignancy is related to tumor microenvironment changes which affect the stiffness values recorded by elastography [45]. This further suggests that elastography measurements may be useful to identify metastatic disease. Shear wave elastography relies on generation of ‘‘adequate” shear waves at the depth of the lesion and therefore for relatively deeper lesions (greater than 4–4.5 cm) a technique has been proposed to determine adequacy of shear waves based on a quality factor [32,46]. This quality factor is based on signal to noise ratio of multiple echoes and induced displacement after ARFI to identify shear wave velocities that may have been mis-encoded as lower value (<4.5 m/s) suggestive of benign nature [46]. Such objective measures to determine adequate implementation of elastography techniques will increase the diagnostic confidence in characterizing breast lesions. As with other ultrasound techniques migrating from 2D toward 3D assessments, applications of 3D SWE for breast lesion characterization are also being evaluated. Two such studies summarized here have reported differently while comparing the performance of 2D and 3D elastography. In a study involving 146 women (163 breast lesions; 48 malignant; 115 benign), the performance of 2D and 3D SWE were compared [47]. For 3D SWE, mean value of Emax
186
J.R. Eisenbrey et al. / Ultrasonics 70 (2016) 183–190
Fig. 2. Example SWE and grayscale images from a benign fibroadenoma (a) and invasive ductal carcinoma (b) obtained using SuperSonic Imagine scanner. Note, the invasive ductal carcinoma showing high stiffness values (>140 kPa) relative to benign fibroadenoma (<50 kPa).
(maximum elasticity value in kPa) of the lesion computed in all three orthogonal planes was compared with the mean value of Emax of the lesion obtained with 2D SWE. Qualitative comparisons between 2D and 3D SWE were based on visual color scale and elasticity homogeneity within the lesion and surrounding tissue. This study concluded that the performance of 2D SWE was better than 3D SWE. Another such study involving 134 women (144 breast lesions; 67 malignant, 77 benign) had contrasting findings and reported that 2D and 3D SWE performed equally in distinguishing benign and malignant lesions [48]. But the study also noted that quantitative elasticity values with 3D approach were greater than 2D SWE for all masses, and this may necessitate titrating cut-off values for 3D SWE. Overall, qualitative and quantitative assessments for comparing 2D and 3D SWE are based on parameters derived for 2D SWE and in future, 3D SWE specific quantitative
parameters as well as research in artifact and noise reduction techniques for 3D SWE acquisitions may further assist in breast lesion characterization. In a relatively recent study, the performance of 3D SWE was compared with 3D ultrasound and dynamic contrast-enhanced magnetic resonance imaging for imaging and treatment monitoring in patients receiving neoadjuvant chemotherapy [49]. In that study, the feasibility of performing 3D SWE was investigated in 23 patients whereas in 10 patients the authors studied changes in lesion stiffness following treatment. The results showed that 3D SWE may be useful to identify early responders to neoadjuvant chemotherapy based on stiffness distribution and changes. Another 3D elastography technique involving compression of the breast examined characterizing breast lesions based on a nonlinear wave propagation parameter [50] derived from curve-
J.R. Eisenbrey et al. / Ultrasonics 70 (2016) 183–190
187
Fig. 3. Example image showing images of a 3D subharmonic imaging breast exam at baseline (a) and 9 s after contrast injection (b), showing enhancement within a complex cyst with apocrine metaplasia.
fitting; the strain differences between the lesion and the surrounding tissues are plotted for different compression levels and the resulting curve-fitting parameters are used for characterization [51]. In a pilot study of 10 women using this approach, the group showed an order of magnitude increase in this nonlinear wave propagation parameter in malignant relative to benign masses, while size and traditional strain ratio values showed only a 2X increase [51]. A follow up study from the same group extended the type of curve-fitting and number of parameters derived from curve-fitting and the use of segmented 3D elastograms to characterize breast lesions [52]. Given that both these studies were performed with a limited sample size of 10 patients, evaluation of
this technique in a larger cohort will ultimately determine clinical relevance.
6. Contrast-enhanced ultrasound Ultrasound contrast agents are comprised of gas microbubbles encapsulated by an outer shell for stability [53]. These microbubbles are traditionally injected intravenously and due to their size (<8 lm) restricted to the vascular space [54]. Insonated microbubbles are also highly nonlinear (producing echoes spanning the subharmonic response through ultra-harmonic frequency
188
J.R. Eisenbrey et al. / Ultrasonics 70 (2016) 183–190
components), which enable filtering approaches for separating microbubble echoes from surrounding tissue during imaging [53]. The safety profiles of these agents is also well established (a retrospective study of FDA approved ultrasound contrast agents used in echocardiography showed a severe reaction rate of 0.01% over 78,383 doses [55]), making it the safest of all contrast media. The role of ultrasound contrast agents is expected to further increase with the advancement of molecularly-targeted agents. Using this approach, targeting ligands are conjugated to the microbubble shell, thereby increasing affinity of the microbubbles at a specific disease site. Recently, an experimental agent targeted to vascular endothelial growth factors frequently expressed in tumors has been used in a phase 1 clinical trial and shown promise for breast cancer characterization and detection [56]. The use of contrast-enhanced ultrasound (CEUS) has been explored extensively for breast imaging as differences in vascularity can be used for characterizing masses. Early CEUS in breast imaging used power and color Doppler techniques, where investigators demonstrated improved signal strength and visualization of tumor vessel morphology using a subjective scoring system by blinded radiologists [57]. However, the ability to characterize masses based on contrast-enhanced Doppler imaging resulted in mixed results, primarily due to an inability discriminate between vascular characteristics of malignant and benign lesions [58,59]. More success has been reported using nonlinear harmonic contrast imaging packages in which pulse or amplitude modulations are used to separate nonlinear microbubble signals at the second harmonic while suppressing linear tissue signal. For example, Wan et al. looked at 91 breast masses with CEUS and found a 91% area under the ROC curve for qualitative interpretation and 77% for quantitative analysis [60]. In a more recently published metaanalysis of CEUS exams using commercial harmonic imaging packages reviewed 16 studies (with 957 lesions) and found that newer generation ultrasound contrast agents (Optison, GE Healthcare, Princeton, NJ; Definity, Lantheus Medical Imaging, N Billerica,
MA; Sonazoid, GE Healthcare, Oslo Norway; SonoVue, Bracco Imaging, Milan, Italy), significantly improved diagnostic precision compared to an older microbubble and that overall sensitivity and specificity were 86% and 79%, respectively [61]. This is a significant improvement in the specificity of mammography alone and may help reduce false positive rates (reported as low as 65% [5,6]). As tissue itself generates higher harmonic signals [62], our group and others have proposed using subharmonic imaging to better isolate microbubble signal (and thus improve the ability to visualize tumor vascularity). The feasibility of breast subharmonic imaging was first performed in a pilot study with 16 breast masses [63]. Cumulative maximum intensity projections of these exams resulted in an area under the ROC curve of 90%, showing significant improvement relative to mammography (p = 0.031) [64]. Quantitative analysis of blood flow parameters from these exams also demonstrated a significant difference between benign and malignant masses (p = 0.0014) [65]. More recently, subharmonic imaging has been implemented on volumetric probes for a multicenter breast imaging trial and preliminary qualitative analysis has shown that vascular heterogeneity can be quantified and used to characterize masses as benign or malignant [66,67]. An example from this ongoing study is provided in Fig. 3, showing images of a 3D subharmonic imaging breast exam at baseline (a) and 9 s after contrast injection (b), showing enhancement within a complex cyst with apocrine metaplasia. Imaging shows subharmonic signal from three orthogonal planes (grayscale) and the rendered volume in gold. Bifurcated vessels between the cystic components are shown (white arrows), while the cystic regions of the mass remain unenhanced. Imaging was performed on a modified Logiq9 scanner with 4C probe (GE Healthcare, Milwaukee, WI) transmitting 4 cycle pusles at 5.8 MHz and receiving at 2.9 MHz. Lymphosonography is a second area in which CEUS is well poised to make significant contributions to breast imaging. Lymphatic channels in tissue provide drainage for interstitial fluid and the transportation of cancer cells through this drainage are a
Fig. 4. Example image showing ultrasound enhancement (nonlinear signal in gold) of a lymphatic channel (white arrows) and lymph node (circled) 5 min after intradermal injection of ultrasound contrast. Imaging was performed using an Acuson S3000 Helix scanner with a 9L4 probe (Siemens Medical Solutions, Mountain View, CA). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
J.R. Eisenbrey et al. / Ultrasonics 70 (2016) 183–190
primary mechanism of metastasis from solid tumors. Identification and localization of the first line of lymph nodes in a tumoraldrainage path (termed sentinel lymph nodes) are important for tumor staging. Traditionally this is performed using either the injection of lymphatic dyes and visually tracing lymphatic channels within an open surgical bed (a highly invasive approach), or by the injection of radiotracer dyes (lymphoscintigraphy) [68]. However, it has been shown in a swine melanoma model that intradermal injection of ultrasound contrast agents enables realtime visualization of lymphatic channels and sentinel lymph nodes, and that this approach is superior to lymphoscintigraphy (in which a radiolabeled dye is injected subcutaneously and traced to the sentinel lymph node using a gamma probe or single photon emission computed tomography) [69]. Initial studies of lymphosonography have been reported for the identification of sentinel nodes in breast cancer patients [70,71]. Our group is currently involved in a Phase 1 trial using intradermal breast injections of the ultrasound contrast agent Sonazoid (GE Healthcare, Oslo, Norway) for lymph node detection. An example of this ongoing work is shown in Fig. 4. Scanning was performed using a 9L4 probe on a Siemens S3000 Scanner (Siemens Medical Solutions, Mountain View, CA) in dual mode with cadence pulse sequencing (gold signal) and B-mode (grayscale) imaging. Five minutes post injection of 1 ml of Sonazoid, contrast is observed within the lymphatic channel (white arrows) and in the lymph node (white circle) in the left axillary region. It is expected that this technique will aid in sentinel node biopsy in future breast cancer surgeries. 7. Conclusions Breast imaging and breast cancer screening remain important areas within the field of radiology. With resolutions now as low 200 lm, automated volumetric scanning, and refined nonlinear processing approaches, commercially available ultrasound systems have demonstrated significant clinical benefit for characterizing breast masses or for breast cancer screening in the absence of mammography. Emerging technological developments such as elastography and contrast-enhanced ultrasound are expected to further expand this role. No other imaging modality currently combines the lack of ionizing radiation, portability, temporal resolution, and cost-saving advantages that ultrasound offers. These attributes are expected to enable ultrasound to continue to expand as a clinical tool for breast imaging. References [1] R. Siegel, K. Miller, A. Jemal, Cancer statistics, 2016, CA Cancer J. Clin. 66 (1) (2016) 7–30. [2] American Cancer Society, Detailed guide to breast cancer staging, American Cancer Society, (Online). (accessed April 6th, 2016). [3] A. Siu, Screening for breast cancer: U.S. preventive services task force, Ann. Intern. Med. 164 (4) (2016) 279–296. [4] S. Zervoudis, G. Iatrakis, E. Tomara, Main controversies in breast cancer, World J. Clin. Oncol. 10 (5) (2014) 359–373. [5] D. Kopans, Breast Imaging, Lippincott Williams & Wilkins, Philadelphia, 2007. [6] H. Zonderland, E. Coerkamp, J. Hermans, et al., Diagnosis of breast cancer: contribution of US as an adjunct, Radiology 213 (1999) 413–422. [7] A. Stravos, Breast Ultrasound, Lippincott Williams & Wilkins, Philadelphia, 2004. [8] K. Taylor, C. Meritt, C. Piccoli, et al., Ultrasound as a complement to mammography and breast examination to characterize breast masses, Ultrasound Med. Biol. 28 (1) (2002) 19–26. [9] I. Ladabaum, S. Panda, C. Daft, Microfabricated ultrasonic transducer array for 3-D imaging and method of operating the same, 2009 Patent U.S. Patent 7,618,373. [10] B. Diarra, M. Robini, P. Tortoli, C. Cachard, H. Liebgott, Design of optimal 2-D nongrid sparse arrays for medical ultrasound, IEEE Trans. Biomed. Eng. 60 (11) (2013) 3093–3102. [11] W. Liu, J. Zagzebski, T. Hall, et al., Acoustic backscatter and effective scatterer size estimates using a 2D CMUT transducer, Phys. Med. Biol. 53 (15) (2008) 4169–4183.
189
[12] S. Biswas, K. Ananthasubramaniam, Clinical utility of three-dimensional echocardiography for the evaluation of ventricular function, Cardiol. Rev. 21 (4) (2013) 184–195. [13] C. Morbach, B. Lin, L. Sugeng, Clinical application of three-dimensional echocardiograph, Prog. Cardiovasc. Dis. 57 (1) (2014) 19–31. [14] W. Berg, J. Blume, J. Cormack, et al., Combined screening with ultrasound and mammography vs mammography alone in women at elevated risk of breast cancer, JAMA 299 (2008) 2151–2163. [15] W. Berg, A. Bandos, E. Mendelson, et al., Ultrasound as the primary screening test for breast cancer: analysis from ACRIN 6666, J. Natl Cancer Inst. 108 (4) (2015). [16] L. Chen, Y. Chen, X. Diao, et al., Comparative study of automated breast 3-D ultrasound and handheld B-mode ultrasound for differentiation of benign and malignant breast masses, Ultrasound Med. Biol. 39 (10) (2013) 1735–1742. [17] R. Brem, L. Tabár, S. Duffy, et al., Assessing improvement in detection of breast cancer with three-dimensional automated breast US in women with dense breast tissue: the SomoInsight Study, Radiology 274 (3) (2015) 663–673. [18] S. Jeh, S. Kim, J. Choi, et al., Comparison of automated breast ultrasonography to handheld ultrasonography in detecting and diagnosing breast lesions, Acta Radiol. 57 (2) (2016) 162–169. [19] J. Kwak, E. Kim, J. You, et al., Variable breast conditions: comparison of conventional and real-time compound ultrasonography, J. Ultrasound Med. 23 (2004) 85–96. [20] W. Berg, J. Blume, J. Cormack, et al., Lesion detection and characterization in a breast US phantom: results of the ACRIN 6666 investigators, Radiology 239 (2006) 693–702. [21] B. Mesurolle, T. Helou, M. El-Khoury, Tissue harmonic imaging, frequency compound imaging, and conventional imaging: use and benefit in breast sonography, J. Ultrasound Med. 26 (2007) 1041–1051. [22] J. Cha, W. Moon, N. Cho, et al., Characterization of benign and malignant Solid breast masses: comparison of conventional US and tissue harmonic imaging, Radiology 242 (2007) 63–69. [23] R. Hooley, L. Scoutt, L. Philpotts, Breast ultrasonography: state of the art, Radiology 268 (3) (2013) 642–659. [24] P. Machado, J. Eisenbrey, B. Cavanaugh, F. Forsberg, New image processing technique for evaluating breast microcalcifications: a comparative study, J. Ultrasound Med. 31 (6) (2012) 885–893. [25] R. Tan, Y. Xiao, Q. Tang, et al., The diagnostic value of micropure imaging in breast suspicious microcalcification, Acad. Radiol. 22 (11) (2015) 1338–1343. [26] P. Machado, J. Eisenbrey, B. Cavanaugh, F. Forsberg, Microcalcifications versus artifacts: initial evaluation of a new ultrasound image processing technique to identify breast microcalcifications in a screening population, Ultrasound Med. Biol. 40 (9) (2014) 2321–2324. [27] E. Ueno, E. Tohno, S. Soeda, Y. Asaoka, K. Itoh, J. Bamber, M. Blaszçzyk, J. Davey, J. Mckinna, Dynamic tests in real-time breast echography, Ultrasound Med. Biol. 14 (1988) 53–57. [28] K. Parker, S. Huang, R. Musulin, R. Lerner, Tissue response to mechanical vibrations for ‘‘sonoelasticity imaging”, Ultrasound Med. Biol. 16 (1990) 241– 246. [29] J. Ophir, I. Cespedes, H. Ponnekanti, Y. Yaszdi, X. Li, Elastography: a quantitative method for imaging the elasticity of biological tissues, Ultrason. Imaging 13 (1991) 111–134. [30] T. Shiina, K. Nightingale, M. Palmeri, T. Hall, J. Bamber, R. Barr, L. Castera, B. Choi, Y. Chou, D. Cosgrove, C. Dietrich, H. Ding, D. Amy, A. Farrokh, G. Ferraioli, C. Filice, M. Friedrich-Rust, K. Nakashima, F. Schafer, I. Sporea, S. Suzuki, S. Wilson, M. Kudo, WFUMB guidelines and recommendations for clinical use of ultrasound elastography: Part 1: Basic principles and terminology, Ultrasound Med. Biol. 41 (2015) 1126–1147. [31] R. Barr, K. Nakashima, D. Amy, D. Cosgrove, A. Farrokh, F. Schafer, J. Bamber, L. Castera, B. Choi, Y. Chou, C. Dietrich, H. Ding, G. Ferraioli, C. Filice, M. FriedrichRust, T. Hall, K. Nightingale, M. Palmeri, T. Shiina, S. Suzuki, I. Sporea, S. Wilson, M. Kudo, WFUMB guidelines and recommendations for clinical use of ultrasound elastography: Part 2: Breast, Ultrasound Med. Biol. 41 (2015) 1148–1160. [32] R. Barr, Sonographic breast elastography, J. Ultrasound Med. 31 (2012) 773– 783. [33] R. Barr, Elastography in clinical practice, Radiol. Clin. North Am. 52 (2014) 1145–1162. [34] B. Garra, E. Cespedes, J. Ophir, S. Spratt, R. Zuurbier, C. Magnant, M. Pennanen, Elastography of breast lesions: initial clinical results, Radiology 202 (1997) 79–86. [35] A. Itoh, E. Ueno, E. Tohno, H. Kamma, H. Takahashi, T. Shiina, M. Yamakawa, T. Matsumura, Breast disease: clinical application of US elastography for diagnosis, Radiology 239 (2006) 341–350. [36] E. Ueno, T. Umemoto, H. Bando, E. Tohno, K. Waki, T. Matsumura, New quantitative method in breast elastography: fat lesion ratio (FLR), in: Proceedings of the Radiological Society of North America Scientific, 2007. [37] A. Athanasiou, A. Tardivon, M. Tanter, B. Sigal-Zafrani, J. Bercoff, T. Deffieux, J. Gennisson, M. Fink, S. Neuenschwander, Breast lesions: quantitative elastography with supersonic shear imaging–preliminary results, Radiology 256 (2010) 297–303. [38] R. Barr, Real-time ultrasound elasticity of the breast:initial clinical results, Ultrasound Q. 26 (2010) 61–66. [39] R. Barr, S. Destounis, L. Lackey II, W. Svensson, C. Balleyguier, C. Smith, Evaluation of breast lesions using sonographic elasticity imaging, J. Ultrasound Med. 31 (2012) 281–287.
190
J.R. Eisenbrey et al. / Ultrasonics 70 (2016) 183–190
[40] W. Berg, D. Cosgrove, C. Doré, F. Schäfer, W. Svensson, R. Hooley, R. Ohlinger, E. Mendelson, C. Balu-Maestro, M. Locatelli, C. Tourasse, B. Cavanaugh, V. Juhan, A. Stavros, A. Tardivon, J. Gay, J. Henry, C. Cohen-Bacrie, B. Investigators, Shearwave elastography improves the specificity of breast US: the BE1 multinational study of 939 masses, Radiology 262 (2012) 435–449. [41] J. Chang, J. Won, K. Lee, I. Park, A. Yi, W. Moon, Comparison of shear-wave and strain ultrasound elastography in the differentiation of benign and malignant breast lesions, Am. J. Roentgenol. 201 (2013) 347–356. [42] J. Youk, E. Son, H. Gweon, H. Kim, Y. Park, J. Kim, Comparison of strain and shear wave elastography for the differentiation of benign from malignant breast lesions, combined with B-mode ultrasonography: qualitative and quantitative assessments, Ultrasound Med. Biol. 40 (2014) 2336–2344. [43] W. Mu, W. Zhong, J. Yao, L. Li, Y. Peng, Y. Wang, L. Liu, Y. Xiao, S. Liu, C. Wu, Y. Jiang, S. Parajuly, P. Xu, Y. Hao, J. Li, B. Luo, H. Zhi, Ultrasonic Elastography Research Based on a Multicenter Study: Adding Strain Ratio after 5-Point Scoring Evaluation or Not, PLos One, 2016. [44] M. Hayashi, Y. Yamamoto, A. Sueta, M. Tomiguchi, M. Yamamoto-Ibusuki, T. Kawasoe, A. Hamada, H. Iwase, Associations between elastography findings and clinicopathological factors in breast cancer, Medicine 94 (2015) 1–9. [45] Y. Hao, X. Guo, B. Ma, L. Zhu, L. Liu, et al., Relationship between ultrasound elastography and myofibroblast distribution in breast cancer and its clinical significance, Scientific Reports, 2016. [46] R. Barr, Z. Zhang, Shear-wave elastography of the breast: value of a quality measure and comparison with strain elastography, Radiology 275 (2015) 45– 53. [47] J. Youk, H. Gweon, E. Son, J. Chung, J. Kim, E. Kim, Three-dimensional shearwave elastography for differentiating benign and malignant breast lesions: comparison with two-dimensional shear-wave elastography, Eur. Radiol. 23 (2013) 1519–1527. [48] S. Lee, J. Chang, W. Kim, M. Bae, N. Cho, A. Yi, H. Koo, S. Kim, J. Kim, W. Moon, Differentiation of benign from malignant solid breast masses: comparison of two-dimensional and three-dimensional shear-wave elastography, Eur. Radiol. 23 (2013) 1015–1026. [49] A. Athanasiou, H. Latorre-Ossa, A. Criton, A. Tardivon, J. Gennisson, M. Tanter, Feasibility of Imaging and Treatment Monitoring of Breast Lesions with ThreeDimensional Shear Wave Elastography, Ultraschall in Med, 2016. [50] P. Lewin, Nonlinear acoustics in ultrasound metrology and other selected applications, Phys. Procedia 3 (1) (2010) 17–23. [51] A. Sayed, G. Layne, J. Abraham, O. Mukdadi, Nonlinear characterization of breast cancer using multi-compression 3D ultrasound elastography in vivo, Ultrasonics 53 (2013) 979–991. [52] A. Sayed, G. Layne, J. Abraham, O. Mukdadi, 3-D visualization and non-linear tissue classification of breast tumors using ultrasound elastography in vivo, Ultrasound Med. Biol. 40 (2014) 1490–1502. [53] B. Goldberg, J. Raichlen, F. Forsberg, Ultrasound Contrast Agents: Basic Principles and Clinical Applications, second ed., Martin Dunitz Ltd, London, 2001. [54] J. Correas, L. Bridal, A. Lesavre, et al., Ultrasound contrast agents: properties, principles of action, tolerance, and artifacts, Eur. Radiol. 11 (2001) 1316–1328.
[55] K. Wei, S. Mulvagh, L. Carson, et al., The safety of Definity and Optison for ultrasound image enhancement: a retrospective analysis of 78,383 administered contrast doses, J. Am. Soc. Echocardiogr. 21 (2008) 1202–1206. [56] L. Abou-Elkacem, S. Bachawal, J. Willmann, Ultrasound molecular imaging: moving towards clinical translation, Eur. J. Radiol. 84 (9) (2015) 1685–1693. [57] H. Madjar, H. Prömpeler, C. Del Favero, A new Doppler signal enhancing agent for flow assessment in breast lesions, Eur. J. Ultrasound 12 (2000) 123–130. [58] M. Stuhrmann, R. Aronius, M. Schietzel, Tumor vascularity of breast lesions: potentials and limits of contrast-enhanced Doppler sonography, AJR Roentgenol. 175 (2000) 1585–1589. [59] B. Goldberg, C. Merritt, L. Needleman, et al., Contrast-enhanced ultrasound and mammography for breast cancer detection,”, J. Ultrasound Med. 20 (2001) S95. [60] C. Wan, J. Du, H. Fang, F. Li, L. Wang, Evaluation of breast lesions by contrast enhanced ultrasound: qualitative and quantitative analysis, Eur. J. Radiol. 81 (4) (2012). pp. e444-50. [61] Q. Hu, X. Wang, S. Zhu, et al., Meta-analysis of contrast-enhanced ultrasound for the differentiation of benign and malignant breast lesions, Acta Radiol. 56 (1) (2015) 25–33. [62] P. Shankar, P. Krishna, V. Newhouse, Advantages of subharmonic over second harmonic backscatter for contrast-to-tissue echo enhancement, Ultrasound Med. Biol. 24 (1998) 395–399. [63] F. Forsberg, C. Piccoli, D. Merton, Breast lesions: imaging with contrastenhanced subharmonic US—initial experience, Radiology 244 (2007) 718–726. [64] J. Dave, F. Forsberg, S. Fernandes, Static and dynamic cumulative maximum intensity display mode for subharmonic breast imaging: a comparative study with mammographic and conventional ultrasound techniques, J. Ultrasound Med. 29 (8) (2010) 1177–1185. [65] J. Eisenbrey, J. Dave, D. Merton, et al., Parametric imaging using subharmonic signals from ultrasound contrast agents in patients with breast lesions, J. Ultrasound Med. 30 (2) (2011) 85–92. [66] A. Sridharan, J. Eisenbrey, P. Machado, et al., Quantitative analysis of vascular heterogeneity in breast lesions using contrast-enhanced 3-D harmonic and subharmonic ultrasound imaging, IEEE Trans. Ultrason. Ferroelectr. Freq. Control 62 (3) (2015) 502–510. [67] T. Kim, A. Giuliano, G. Lyman, Lymphatic mapping and sentinel lymph node biopsy in early-stage breast carcinoma, Cancer 106 (1) (2006) 4–26. [68] B.B. Goldberg, D.A. Merton, J.B. Liu, et al., Contrast-enhanced ultrasound imaging of sentinel lymph nodes after peritumoral administration of Sonazoid in a melanoma tumor animal model, J. Ultrasound Med. 4 (30) (2011) 441– 453. [69] S. Rautiainen, M. Sudah, S. Joukainen, et al., Contrast-enhanced ultrasound guided axillary lymph node core biopsy: diagnostic accuracy in preoperative staging of invasive breast cancer, Eur. J. Radiol. 84 (11) (2015) 2130–2136. [70] K. Omoto, H. Matsunaga, N. Take, et al., Sentinel node detection method using contrast-enhanced ultrasonography with sonazoid in breast cancer: preliminary clinical study, Ultrasound Med. Biol. 35 (8) (2009) 1249–1256. [71] J. Eisenbrey, A. Sridharan, P. Machado, H. Zhao, Three-dimensional subharmonic ultrasound imaging in vitro and in vivo, Acad Radiol. 19 (6) (2012) 732–739.
Contents lists available at ScienceDirect
Ultrasonics journal homepage: www.elsevier.com/locate/ultras
Recent technological advancements in breast ultrasound John R. Eisenbrey ⇑, Jaydev K. Dave, Flemming Forsberg Thomas Jefferson University, Department of Radiology, Division of Ultrasound, 132 South 10th St., Philadelphia, PA 19107, United States
a r t i c l e
i n f o
Article history: Received 16 March 2016 Received in revised form 20 April 2016 Accepted 24 April 2016 Available online 25 April 2016 Keywords: Breast ultrasound Automated breast ultrasound Elastography Contrast-enhanced ultrasound
a b s t r a c t Ultrasound is becoming increasingly common as an imaging tool for the detection and characterization of breast tumors. This paper provides an overview of recent technological advancements, especially those that may have an impact in clinical applications in the field of breast ultrasound in the near future. These advancements include close to 100% fractional bandwidth high frequency (5–18 MHz) 2D and 3D arrays, automated breast imaging systems to minimize the operator dependence and advanced processing techniques, such as those used for detection of microcalcifications. In addition, elastography and contrast-enhanced ultrasound examinations that are expected to further enhance the clinical importance of ultrasound based breast tumor screening are briefly reviewed. These techniques have shown initial promise in clinical trials and may translate to more comprehensive clinical adoption in the future. Ó 2016 Elsevier B.V. All rights reserved.
Contents 1. 2. 3. 4. 5. 6. 7.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 2D and 3D transducer technologies . . . . . . . . Whole breast ultrasound screening . . . . . . . . Image acquisition and processing techniques Elastography . . . . . . . . . . . . . . . . . . . . . . . . . . . Contrast-enhanced ultrasound . . . . . . . . . . . . Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
1. Introduction Breast cancer is the single most commonly occurring cancer in women in the United States with over 249,000 cases, and second most common cause of cancer death in women (41,152 deaths expected in 2016) [1]. Breast cancer screening is an important step for the early diagnosis of malignancy as 5 year survival rates vary dramatically by stage of initial diagnosis (as high as 100% at stage I (see [2] for staging criteria) and roughly 25% when diagnosed at stage IV) [1]. X-ray mammography is currently the primary imaging screening tool and recommended in the United States for women 40 years and older [3], although considerable controversy currently surrounds the appropriate age to begin screening [4]. Despite its advantages mammography suffers from poor overall ⇑ Corresponding author. E-mail address: [email protected] (J.R. Eisenbrey). http://dx.doi.org/10.1016/j.ultras.2016.04.021 0041-624X/Ó 2016 Elsevier B.V. All rights reserved.
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
183 184 184 184 185 187 189 189
specificity, resulting in false-positive rates of 65–90% [5,6]. Additionally, mammography is often not adequate in the 20–50% of patients with dense breasts [5]. Ultrasound may overcome both of these limitations. Breast density does not inhibit ultrasound waves in the breast making breast ultrasound imaging a useful technique in these women [7]. Breast ultrasound has also been shown to be a useful adjunct for the characterization of masses post-mammography using an ultrasound-based Breast Imaging Reporting and Data Systems (BIRAD) scoring system. Breast ultrasound post mammography has shown to result in a down staging of roughly 20–40% of BIRADS 4A cases and of roughly 15–25% of BIRADS 3 cases, thus limiting the number of patients requiring biopsy [7]. Unnecessary biopsies are undesirable due to patient discomfort and anxiety, risk of infection, and cost of the procedure. When biopsy is still required, ultrasound is also routinely used for image guided breast mass biopsies [7].
184
J.R. Eisenbrey et al. / Ultrasonics 70 (2016) 183–190
While ultrasound is already playing an important clinical role in breast imaging, overall specificities are still relatively low (around 66% based on a retrospective review of 761 cases using only Bmode imaging [8]). Hence, numerous techniques are currently in development to improve ultrasound’s prominence in this application. This paper will serve as an overview of emerging concepts in the field, including high frequency imaging and processing approaches, automated volumetric breast ultrasound systems, breast elastography, and contrast-enhanced breast ultrasound. 2. 2D and 3D transducer technologies High frequency, broad bandwidth transducers can produce excellent detail resolution at the relatively shallow depths (typically < 4 cm) encountered in breast imaging. Currently available commercial systems use linear arrays operating around 10– 14 MHz with close to 100% bandwidth ranging from 5 to 18 MHz. With a typical (lateral) resolution close to 2k, a 10 MHz transmit would yield resolution of approximately 300 lm. Thus, such arrays are capable of providing superior resolution at multiple depths by selecting the best possible compromise between penetration depth, which is dependent on attenuation of the pressure amplitude probing the tissue, and the wavelength dependent resolution. Examples of bandwidths in excess of 150% with center frequencies of 9–15 MHz have been reported using Polyvinylidene Fluoride (PVDF) as well as capacitive microfabricated ultrasound transducer (cMUT) probes [9–11]. A major goal in transducer development has been the construction of electronic volumetric arrays, which would allow for simultaneous beam-steering in the axial, lateral and azimuthal directions and the acquisition of full volumetric data sets (focused in all three dimensions) [10–13]. Moreover, volumetric arrays would permit true vector velocity flow imaging, since a volume data set allows flow to be interrogated at multiple angles, which could be combined to determine the true 3D flow vector. The challenge in construction of volumetric arrays has always been the large number of elements (10,000+) and, while multiplexers in the transducer handle have limited the number of wires necessary for use it will not be trivial to construct volumetric arrays operating at the higher frequencies required in breast imaging. Nonetheless, commercial volumetric phased as well as linear arrays for use in echocardiography are now available from several manufacturers [12,13]. These matrix arrays can produce real time 3D imaging (over 20 volumes per second), but they operate at frequencies too low to be suitable for breast imaging (i.e., <7 MHz; resolution of approximately 500 lm). 3. Whole breast ultrasound screening Whole breast ultrasound screening studies to date have focused on women with elevated breast cancer risk and/or dense breast tissue [14,15]. When comparing mammography alone to screening with ultrasound and mammography the largest study to date (the ACRIN 6666 trial) involving over 2800 women found an additional 1.1 to 7.2 cancers per 1000 high-risk women; albeit with a marked increase in the number of false positives [14]. This group of investigators also compared ultrasound screening alone to conventional mammographic screening and concluded that cancer detection rates were comparable [15]. As before, false positives were more common with whole breast ultrasound screening. However, screening examinations can be time consuming and hand held transducers are by definition limited to a 2D field of view. These exams may take up to 30 min compared to roughly 10 min for a mammogram, although with less patient discomfort and no associated radiation. As an alternative, large field of view
systems capable of scanning an entire breast using very large linear arrays (14–15 cm in length) have been developed for screening purposes [16–18]. Automated 3D whole breast ultrasound systems allow the technologist to acquire high quality images (resolutions < 300 lm) in the standard mammographic views across three planes and have been shown to achieve the same diagnostic accuracy (>87%) as hand held ultrasound [16,17]. The FDA approved automated breast ultrasound in 2012 as an adjunct screening modality to mammography in asymptomatic women with dense breast tissue (breast density scores of 3–4, based on a mammography-indicated density of over 51%) for whom screening mammography findings are normal or benign. With depiction of the coronal plane, mass margins, shape, spiculations, and distortion associated with tissue retraction are visible [16–18].
4. Image acquisition and processing techniques Two image acquisition techniques – compound imaging and tissue harmonic imaging (THI) – have become the bread and butter processing techniques of clinical breast imaging [19–23]. In spatial compound imaging several (typically 3–9) ultrasound images are acquired at different angles of insonation and averaged to produce a single image. Averaging reduces speckle and improves the delineation of lateral borders, which results in increased conspicuity of low-contrast lesions, enhanced delineation of capsular margins and ducts, and better overall image quality when using spatial compounding compared to conventional imaging [19,20,23]. Similarly, frequency compounding uses multiple filters to create images from different received frequency bands. These images are subsequently averaged to suppress speckle, increase contrast resolution and improve penetration [21]. Frequency compounding is also less susceptible to the depth-related issues that affect spatial compounding. In THI mode, the scanner is configured to receive echoes at double the transmit (or fundamental) frequency. By optimizing the receive filters for the weak, nonlinear tissue signal components that arise once sound waves propagate some frequency dependent distance into the tissue, it is possible to reduce echoes from clutter and side lobes at the fundamental frequency, while improving contrast resolution and border delineation [22,23]. The grayscale contrast between glandular or fatty tissue and breast lesions is improved with THI compared with conventional breast ultrasound [22]. The simultaneous use of spatial compounding and THI increases overall image quality and lesion conspicuity, and is now recommended in routine clinical practice [23]. MicroPure (Toshiba America Medical Systems, Tustin, CA) is a new commercial ultrasound image processing technique that post-processes grayscale ultrasound images in order to improve the visualization of breast microcalcifications [24–26]. This software uses a filter technique, where each pixel is compared to the average brightness of the surrounding area, to detect locations where there is a characteristic change from surrounding area [24]. As the focus of MicroPure imaging is to detect microcalcifications in breast tissue, the filter kernel is optimized in the horizontal direction to detect only isolated points with higher brightness compared to the surrounding breast tissue [24–26]. An example of this technique being used to improve ultrasound visualization of microcalcifications in the breast is provided in Fig. 1. Independent investigations have established that MicroPure can detect more microcalcifications than standard ultrasound in women presenting with microcalcifications seen on mammography [24,25]. However, this method is not appropriate for a screening population as it does not possess the same level of sensitivity as mammography, and should be used in more focused applications such as for guiding a biopsy [26].
J.R. Eisenbrey et al. / Ultrasonics 70 (2016) 183–190
185
Fig. 1. Example imaging of MicroPure imaging being used to visualize breast microcalcifications. Dual imaging is used in which grayscale B-mode ultrasound is displayed in left side of the image, and MicroPure is displayed in blue on the right. Calcifications are identified by the brighter white echoes overlain on the blue image. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
5. Elastography Elastography based techniques convey information about stiffness of a lesion under consideration, with stiffer tissue generally being indicative of malignancy. Since early published literature on the use of ultrasound for elastography [27–29] several singleand multi- center studies have been conducted to evaluate the usefulness of elastography to characterize breast lesions, and the most recent guidelines for clinical use of elastography were published in 2015 by World Federation of Ultrasound in Medicine and Biology (WFUMB) and recommend elastography for the characterization of solid breast masses [30,31]. The overall principle of elastography depends on introducing mechanical excitation within a region of interest and measuring the induced disturbance. The induced disturbance is measured either as displacement within the field of view and this is referred to as strain elastography (SE) or as the velocity of an induced shear wave and this is referred to as shear wave elastography (SWE) [32,33,30,31]. Both SE and SWE are used for characterizing breast lesions. From the technological perspective for SE of breast lesions, the induced disturbance is compression either achieved manually or using a subject’s respiratory or cardiac pulsations, or via Acoustic Radiation Force Imaging (ARFI) wherein the ultrasound transducer is used to initiate a pulse resulting in disturbance. For SWE of breast lesions, again ARFI is used but the speed of the induced shear wave in the lesion is measured (SWE with manual compression is not performed). The measured shear wave speed (in m/s) can also be expressed as stiffness (in kPa) after conversion (assuming a Young’s modulus of 0.5) as shown in Fig. 2. The characterization of breast lesions as benign or malignant based on SE and SWE is based on the principle that benign lesions are relatively easily deformable and therefore, characterized by low stiffness values as compared to malignant lesions that exhibit increased stiffness values. The elastography images are interpreted based on qualitative, semi-quantitative and quantitative assessments [34–37]. In brief, these assessments involve characterization of lesions based on a lesion-stiffness dependent color coded images derived from SE or SWE and measurements based on grayscale and corresponding elastography images including ratios of lesion size and strain measurements in fat and lesion. Based on these assessments it has been shown that a major advantage of utilizing elastography for breast lesion characterization is the improvement in
specificity up to 20% (relative to B-mode ultrasound) to rule out biopsy specifically to distinguish lesions categorized as BI-RADS category 3 or 4A [35,37–40]. Two recent studies have shown that the performance of SE and SWE for differentiation between benign and malignant breast lesions is similar with equivalent areas under their receiver operating characteristic curves (AUC > 0.9) [41,42]. Another multi-center study involving subjects compared two SE techniques, namely, a qualitative scoring system for elastography images and assessment of strain ratios, and concluded that in their study population the diagnostic performance characterized by these two SE techniques was not different (area under the receiver operator characteristic (ROC) curve for both techniques was 0.86). [43]. Such data suggest that following guidelines as recommended by WFUMB [31], both SE and SWE techniques may yield similar performance. Preliminary results also indicate that breast lesion stiffness may be closely associated with lymph node involvement and therefore SE or SWE may also be useful in characterizing the metastatic nature of a malignant lesion [44]. Also, the microenvironment of benign and malignant cells is different and invasiveness associated with malignancy is related to tumor microenvironment changes which affect the stiffness values recorded by elastography [45]. This further suggests that elastography measurements may be useful to identify metastatic disease. Shear wave elastography relies on generation of ‘‘adequate” shear waves at the depth of the lesion and therefore for relatively deeper lesions (greater than 4–4.5 cm) a technique has been proposed to determine adequacy of shear waves based on a quality factor [32,46]. This quality factor is based on signal to noise ratio of multiple echoes and induced displacement after ARFI to identify shear wave velocities that may have been mis-encoded as lower value (<4.5 m/s) suggestive of benign nature [46]. Such objective measures to determine adequate implementation of elastography techniques will increase the diagnostic confidence in characterizing breast lesions. As with other ultrasound techniques migrating from 2D toward 3D assessments, applications of 3D SWE for breast lesion characterization are also being evaluated. Two such studies summarized here have reported differently while comparing the performance of 2D and 3D elastography. In a study involving 146 women (163 breast lesions; 48 malignant; 115 benign), the performance of 2D and 3D SWE were compared [47]. For 3D SWE, mean value of Emax
186
J.R. Eisenbrey et al. / Ultrasonics 70 (2016) 183–190
Fig. 2. Example SWE and grayscale images from a benign fibroadenoma (a) and invasive ductal carcinoma (b) obtained using SuperSonic Imagine scanner. Note, the invasive ductal carcinoma showing high stiffness values (>140 kPa) relative to benign fibroadenoma (<50 kPa).
(maximum elasticity value in kPa) of the lesion computed in all three orthogonal planes was compared with the mean value of Emax of the lesion obtained with 2D SWE. Qualitative comparisons between 2D and 3D SWE were based on visual color scale and elasticity homogeneity within the lesion and surrounding tissue. This study concluded that the performance of 2D SWE was better than 3D SWE. Another such study involving 134 women (144 breast lesions; 67 malignant, 77 benign) had contrasting findings and reported that 2D and 3D SWE performed equally in distinguishing benign and malignant lesions [48]. But the study also noted that quantitative elasticity values with 3D approach were greater than 2D SWE for all masses, and this may necessitate titrating cut-off values for 3D SWE. Overall, qualitative and quantitative assessments for comparing 2D and 3D SWE are based on parameters derived for 2D SWE and in future, 3D SWE specific quantitative
parameters as well as research in artifact and noise reduction techniques for 3D SWE acquisitions may further assist in breast lesion characterization. In a relatively recent study, the performance of 3D SWE was compared with 3D ultrasound and dynamic contrast-enhanced magnetic resonance imaging for imaging and treatment monitoring in patients receiving neoadjuvant chemotherapy [49]. In that study, the feasibility of performing 3D SWE was investigated in 23 patients whereas in 10 patients the authors studied changes in lesion stiffness following treatment. The results showed that 3D SWE may be useful to identify early responders to neoadjuvant chemotherapy based on stiffness distribution and changes. Another 3D elastography technique involving compression of the breast examined characterizing breast lesions based on a nonlinear wave propagation parameter [50] derived from curve-
J.R. Eisenbrey et al. / Ultrasonics 70 (2016) 183–190
187
Fig. 3. Example image showing images of a 3D subharmonic imaging breast exam at baseline (a) and 9 s after contrast injection (b), showing enhancement within a complex cyst with apocrine metaplasia.
fitting; the strain differences between the lesion and the surrounding tissues are plotted for different compression levels and the resulting curve-fitting parameters are used for characterization [51]. In a pilot study of 10 women using this approach, the group showed an order of magnitude increase in this nonlinear wave propagation parameter in malignant relative to benign masses, while size and traditional strain ratio values showed only a 2X increase [51]. A follow up study from the same group extended the type of curve-fitting and number of parameters derived from curve-fitting and the use of segmented 3D elastograms to characterize breast lesions [52]. Given that both these studies were performed with a limited sample size of 10 patients, evaluation of
this technique in a larger cohort will ultimately determine clinical relevance.
6. Contrast-enhanced ultrasound Ultrasound contrast agents are comprised of gas microbubbles encapsulated by an outer shell for stability [53]. These microbubbles are traditionally injected intravenously and due to their size (<8 lm) restricted to the vascular space [54]. Insonated microbubbles are also highly nonlinear (producing echoes spanning the subharmonic response through ultra-harmonic frequency
188
J.R. Eisenbrey et al. / Ultrasonics 70 (2016) 183–190
components), which enable filtering approaches for separating microbubble echoes from surrounding tissue during imaging [53]. The safety profiles of these agents is also well established (a retrospective study of FDA approved ultrasound contrast agents used in echocardiography showed a severe reaction rate of 0.01% over 78,383 doses [55]), making it the safest of all contrast media. The role of ultrasound contrast agents is expected to further increase with the advancement of molecularly-targeted agents. Using this approach, targeting ligands are conjugated to the microbubble shell, thereby increasing affinity of the microbubbles at a specific disease site. Recently, an experimental agent targeted to vascular endothelial growth factors frequently expressed in tumors has been used in a phase 1 clinical trial and shown promise for breast cancer characterization and detection [56]. The use of contrast-enhanced ultrasound (CEUS) has been explored extensively for breast imaging as differences in vascularity can be used for characterizing masses. Early CEUS in breast imaging used power and color Doppler techniques, where investigators demonstrated improved signal strength and visualization of tumor vessel morphology using a subjective scoring system by blinded radiologists [57]. However, the ability to characterize masses based on contrast-enhanced Doppler imaging resulted in mixed results, primarily due to an inability discriminate between vascular characteristics of malignant and benign lesions [58,59]. More success has been reported using nonlinear harmonic contrast imaging packages in which pulse or amplitude modulations are used to separate nonlinear microbubble signals at the second harmonic while suppressing linear tissue signal. For example, Wan et al. looked at 91 breast masses with CEUS and found a 91% area under the ROC curve for qualitative interpretation and 77% for quantitative analysis [60]. In a more recently published metaanalysis of CEUS exams using commercial harmonic imaging packages reviewed 16 studies (with 957 lesions) and found that newer generation ultrasound contrast agents (Optison, GE Healthcare, Princeton, NJ; Definity, Lantheus Medical Imaging, N Billerica,
MA; Sonazoid, GE Healthcare, Oslo Norway; SonoVue, Bracco Imaging, Milan, Italy), significantly improved diagnostic precision compared to an older microbubble and that overall sensitivity and specificity were 86% and 79%, respectively [61]. This is a significant improvement in the specificity of mammography alone and may help reduce false positive rates (reported as low as 65% [5,6]). As tissue itself generates higher harmonic signals [62], our group and others have proposed using subharmonic imaging to better isolate microbubble signal (and thus improve the ability to visualize tumor vascularity). The feasibility of breast subharmonic imaging was first performed in a pilot study with 16 breast masses [63]. Cumulative maximum intensity projections of these exams resulted in an area under the ROC curve of 90%, showing significant improvement relative to mammography (p = 0.031) [64]. Quantitative analysis of blood flow parameters from these exams also demonstrated a significant difference between benign and malignant masses (p = 0.0014) [65]. More recently, subharmonic imaging has been implemented on volumetric probes for a multicenter breast imaging trial and preliminary qualitative analysis has shown that vascular heterogeneity can be quantified and used to characterize masses as benign or malignant [66,67]. An example from this ongoing study is provided in Fig. 3, showing images of a 3D subharmonic imaging breast exam at baseline (a) and 9 s after contrast injection (b), showing enhancement within a complex cyst with apocrine metaplasia. Imaging shows subharmonic signal from three orthogonal planes (grayscale) and the rendered volume in gold. Bifurcated vessels between the cystic components are shown (white arrows), while the cystic regions of the mass remain unenhanced. Imaging was performed on a modified Logiq9 scanner with 4C probe (GE Healthcare, Milwaukee, WI) transmitting 4 cycle pusles at 5.8 MHz and receiving at 2.9 MHz. Lymphosonography is a second area in which CEUS is well poised to make significant contributions to breast imaging. Lymphatic channels in tissue provide drainage for interstitial fluid and the transportation of cancer cells through this drainage are a
Fig. 4. Example image showing ultrasound enhancement (nonlinear signal in gold) of a lymphatic channel (white arrows) and lymph node (circled) 5 min after intradermal injection of ultrasound contrast. Imaging was performed using an Acuson S3000 Helix scanner with a 9L4 probe (Siemens Medical Solutions, Mountain View, CA). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
J.R. Eisenbrey et al. / Ultrasonics 70 (2016) 183–190
primary mechanism of metastasis from solid tumors. Identification and localization of the first line of lymph nodes in a tumoraldrainage path (termed sentinel lymph nodes) are important for tumor staging. Traditionally this is performed using either the injection of lymphatic dyes and visually tracing lymphatic channels within an open surgical bed (a highly invasive approach), or by the injection of radiotracer dyes (lymphoscintigraphy) [68]. However, it has been shown in a swine melanoma model that intradermal injection of ultrasound contrast agents enables realtime visualization of lymphatic channels and sentinel lymph nodes, and that this approach is superior to lymphoscintigraphy (in which a radiolabeled dye is injected subcutaneously and traced to the sentinel lymph node using a gamma probe or single photon emission computed tomography) [69]. Initial studies of lymphosonography have been reported for the identification of sentinel nodes in breast cancer patients [70,71]. Our group is currently involved in a Phase 1 trial using intradermal breast injections of the ultrasound contrast agent Sonazoid (GE Healthcare, Oslo, Norway) for lymph node detection. An example of this ongoing work is shown in Fig. 4. Scanning was performed using a 9L4 probe on a Siemens S3000 Scanner (Siemens Medical Solutions, Mountain View, CA) in dual mode with cadence pulse sequencing (gold signal) and B-mode (grayscale) imaging. Five minutes post injection of 1 ml of Sonazoid, contrast is observed within the lymphatic channel (white arrows) and in the lymph node (white circle) in the left axillary region. It is expected that this technique will aid in sentinel node biopsy in future breast cancer surgeries. 7. Conclusions Breast imaging and breast cancer screening remain important areas within the field of radiology. With resolutions now as low 200 lm, automated volumetric scanning, and refined nonlinear processing approaches, commercially available ultrasound systems have demonstrated significant clinical benefit for characterizing breast masses or for breast cancer screening in the absence of mammography. Emerging technological developments such as elastography and contrast-enhanced ultrasound are expected to further expand this role. No other imaging modality currently combines the lack of ionizing radiation, portability, temporal resolution, and cost-saving advantages that ultrasound offers. These attributes are expected to enable ultrasound to continue to expand as a clinical tool for breast imaging. References [1] R. Siegel, K. Miller, A. Jemal, Cancer statistics, 2016, CA Cancer J. Clin. 66 (1) (2016) 7–30. [2] American Cancer Society, Detailed guide to breast cancer staging, American Cancer Society, (Online).
189
[12] S. Biswas, K. Ananthasubramaniam, Clinical utility of three-dimensional echocardiography for the evaluation of ventricular function, Cardiol. Rev. 21 (4) (2013) 184–195. [13] C. Morbach, B. Lin, L. Sugeng, Clinical application of three-dimensional echocardiograph, Prog. Cardiovasc. Dis. 57 (1) (2014) 19–31. [14] W. Berg, J. Blume, J. Cormack, et al., Combined screening with ultrasound and mammography vs mammography alone in women at elevated risk of breast cancer, JAMA 299 (2008) 2151–2163. [15] W. Berg, A. Bandos, E. Mendelson, et al., Ultrasound as the primary screening test for breast cancer: analysis from ACRIN 6666, J. Natl Cancer Inst. 108 (4) (2015). [16] L. Chen, Y. Chen, X. Diao, et al., Comparative study of automated breast 3-D ultrasound and handheld B-mode ultrasound for differentiation of benign and malignant breast masses, Ultrasound Med. Biol. 39 (10) (2013) 1735–1742. [17] R. Brem, L. Tabár, S. Duffy, et al., Assessing improvement in detection of breast cancer with three-dimensional automated breast US in women with dense breast tissue: the SomoInsight Study, Radiology 274 (3) (2015) 663–673. [18] S. Jeh, S. Kim, J. Choi, et al., Comparison of automated breast ultrasonography to handheld ultrasonography in detecting and diagnosing breast lesions, Acta Radiol. 57 (2) (2016) 162–169. [19] J. Kwak, E. Kim, J. You, et al., Variable breast conditions: comparison of conventional and real-time compound ultrasonography, J. Ultrasound Med. 23 (2004) 85–96. [20] W. Berg, J. Blume, J. Cormack, et al., Lesion detection and characterization in a breast US phantom: results of the ACRIN 6666 investigators, Radiology 239 (2006) 693–702. [21] B. Mesurolle, T. Helou, M. El-Khoury, Tissue harmonic imaging, frequency compound imaging, and conventional imaging: use and benefit in breast sonography, J. Ultrasound Med. 26 (2007) 1041–1051. [22] J. Cha, W. Moon, N. Cho, et al., Characterization of benign and malignant Solid breast masses: comparison of conventional US and tissue harmonic imaging, Radiology 242 (2007) 63–69. [23] R. Hooley, L. Scoutt, L. Philpotts, Breast ultrasonography: state of the art, Radiology 268 (3) (2013) 642–659. [24] P. Machado, J. Eisenbrey, B. Cavanaugh, F. Forsberg, New image processing technique for evaluating breast microcalcifications: a comparative study, J. Ultrasound Med. 31 (6) (2012) 885–893. [25] R. Tan, Y. Xiao, Q. Tang, et al., The diagnostic value of micropure imaging in breast suspicious microcalcification, Acad. Radiol. 22 (11) (2015) 1338–1343. [26] P. Machado, J. Eisenbrey, B. Cavanaugh, F. Forsberg, Microcalcifications versus artifacts: initial evaluation of a new ultrasound image processing technique to identify breast microcalcifications in a screening population, Ultrasound Med. Biol. 40 (9) (2014) 2321–2324. [27] E. Ueno, E. Tohno, S. Soeda, Y. Asaoka, K. Itoh, J. Bamber, M. Blaszçzyk, J. Davey, J. Mckinna, Dynamic tests in real-time breast echography, Ultrasound Med. Biol. 14 (1988) 53–57. [28] K. Parker, S. Huang, R. Musulin, R. Lerner, Tissue response to mechanical vibrations for ‘‘sonoelasticity imaging”, Ultrasound Med. Biol. 16 (1990) 241– 246. [29] J. Ophir, I. Cespedes, H. Ponnekanti, Y. Yaszdi, X. Li, Elastography: a quantitative method for imaging the elasticity of biological tissues, Ultrason. Imaging 13 (1991) 111–134. [30] T. Shiina, K. Nightingale, M. Palmeri, T. Hall, J. Bamber, R. Barr, L. Castera, B. Choi, Y. Chou, D. Cosgrove, C. Dietrich, H. Ding, D. Amy, A. Farrokh, G. Ferraioli, C. Filice, M. Friedrich-Rust, K. Nakashima, F. Schafer, I. Sporea, S. Suzuki, S. Wilson, M. Kudo, WFUMB guidelines and recommendations for clinical use of ultrasound elastography: Part 1: Basic principles and terminology, Ultrasound Med. Biol. 41 (2015) 1126–1147. [31] R. Barr, K. Nakashima, D. Amy, D. Cosgrove, A. Farrokh, F. Schafer, J. Bamber, L. Castera, B. Choi, Y. Chou, C. Dietrich, H. Ding, G. Ferraioli, C. Filice, M. FriedrichRust, T. Hall, K. Nightingale, M. Palmeri, T. Shiina, S. Suzuki, I. Sporea, S. Wilson, M. Kudo, WFUMB guidelines and recommendations for clinical use of ultrasound elastography: Part 2: Breast, Ultrasound Med. Biol. 41 (2015) 1148–1160. [32] R. Barr, Sonographic breast elastography, J. Ultrasound Med. 31 (2012) 773– 783. [33] R. Barr, Elastography in clinical practice, Radiol. Clin. North Am. 52 (2014) 1145–1162. [34] B. Garra, E. Cespedes, J. Ophir, S. Spratt, R. Zuurbier, C. Magnant, M. Pennanen, Elastography of breast lesions: initial clinical results, Radiology 202 (1997) 79–86. [35] A. Itoh, E. Ueno, E. Tohno, H. Kamma, H. Takahashi, T. Shiina, M. Yamakawa, T. Matsumura, Breast disease: clinical application of US elastography for diagnosis, Radiology 239 (2006) 341–350. [36] E. Ueno, T. Umemoto, H. Bando, E. Tohno, K. Waki, T. Matsumura, New quantitative method in breast elastography: fat lesion ratio (FLR), in: Proceedings of the Radiological Society of North America Scientific, 2007. [37] A. Athanasiou, A. Tardivon, M. Tanter, B. Sigal-Zafrani, J. Bercoff, T. Deffieux, J. Gennisson, M. Fink, S. Neuenschwander, Breast lesions: quantitative elastography with supersonic shear imaging–preliminary results, Radiology 256 (2010) 297–303. [38] R. Barr, Real-time ultrasound elasticity of the breast:initial clinical results, Ultrasound Q. 26 (2010) 61–66. [39] R. Barr, S. Destounis, L. Lackey II, W. Svensson, C. Balleyguier, C. Smith, Evaluation of breast lesions using sonographic elasticity imaging, J. Ultrasound Med. 31 (2012) 281–287.
190
J.R. Eisenbrey et al. / Ultrasonics 70 (2016) 183–190
[40] W. Berg, D. Cosgrove, C. Doré, F. Schäfer, W. Svensson, R. Hooley, R. Ohlinger, E. Mendelson, C. Balu-Maestro, M. Locatelli, C. Tourasse, B. Cavanaugh, V. Juhan, A. Stavros, A. Tardivon, J. Gay, J. Henry, C. Cohen-Bacrie, B. Investigators, Shearwave elastography improves the specificity of breast US: the BE1 multinational study of 939 masses, Radiology 262 (2012) 435–449. [41] J. Chang, J. Won, K. Lee, I. Park, A. Yi, W. Moon, Comparison of shear-wave and strain ultrasound elastography in the differentiation of benign and malignant breast lesions, Am. J. Roentgenol. 201 (2013) 347–356. [42] J. Youk, E. Son, H. Gweon, H. Kim, Y. Park, J. Kim, Comparison of strain and shear wave elastography for the differentiation of benign from malignant breast lesions, combined with B-mode ultrasonography: qualitative and quantitative assessments, Ultrasound Med. Biol. 40 (2014) 2336–2344. [43] W. Mu, W. Zhong, J. Yao, L. Li, Y. Peng, Y. Wang, L. Liu, Y. Xiao, S. Liu, C. Wu, Y. Jiang, S. Parajuly, P. Xu, Y. Hao, J. Li, B. Luo, H. Zhi, Ultrasonic Elastography Research Based on a Multicenter Study: Adding Strain Ratio after 5-Point Scoring Evaluation or Not, PLos One, 2016. [44] M. Hayashi, Y. Yamamoto, A. Sueta, M. Tomiguchi, M. Yamamoto-Ibusuki, T. Kawasoe, A. Hamada, H. Iwase, Associations between elastography findings and clinicopathological factors in breast cancer, Medicine 94 (2015) 1–9. [45] Y. Hao, X. Guo, B. Ma, L. Zhu, L. Liu, et al., Relationship between ultrasound elastography and myofibroblast distribution in breast cancer and its clinical significance, Scientific Reports, 2016. [46] R. Barr, Z. Zhang, Shear-wave elastography of the breast: value of a quality measure and comparison with strain elastography, Radiology 275 (2015) 45– 53. [47] J. Youk, H. Gweon, E. Son, J. Chung, J. Kim, E. Kim, Three-dimensional shearwave elastography for differentiating benign and malignant breast lesions: comparison with two-dimensional shear-wave elastography, Eur. Radiol. 23 (2013) 1519–1527. [48] S. Lee, J. Chang, W. Kim, M. Bae, N. Cho, A. Yi, H. Koo, S. Kim, J. Kim, W. Moon, Differentiation of benign from malignant solid breast masses: comparison of two-dimensional and three-dimensional shear-wave elastography, Eur. Radiol. 23 (2013) 1015–1026. [49] A. Athanasiou, H. Latorre-Ossa, A. Criton, A. Tardivon, J. Gennisson, M. Tanter, Feasibility of Imaging and Treatment Monitoring of Breast Lesions with ThreeDimensional Shear Wave Elastography, Ultraschall in Med, 2016. [50] P. Lewin, Nonlinear acoustics in ultrasound metrology and other selected applications, Phys. Procedia 3 (1) (2010) 17–23. [51] A. Sayed, G. Layne, J. Abraham, O. Mukdadi, Nonlinear characterization of breast cancer using multi-compression 3D ultrasound elastography in vivo, Ultrasonics 53 (2013) 979–991. [52] A. Sayed, G. Layne, J. Abraham, O. Mukdadi, 3-D visualization and non-linear tissue classification of breast tumors using ultrasound elastography in vivo, Ultrasound Med. Biol. 40 (2014) 1490–1502. [53] B. Goldberg, J. Raichlen, F. Forsberg, Ultrasound Contrast Agents: Basic Principles and Clinical Applications, second ed., Martin Dunitz Ltd, London, 2001. [54] J. Correas, L. Bridal, A. Lesavre, et al., Ultrasound contrast agents: properties, principles of action, tolerance, and artifacts, Eur. Radiol. 11 (2001) 1316–1328.
[55] K. Wei, S. Mulvagh, L. Carson, et al., The safety of Definity and Optison for ultrasound image enhancement: a retrospective analysis of 78,383 administered contrast doses, J. Am. Soc. Echocardiogr. 21 (2008) 1202–1206. [56] L. Abou-Elkacem, S. Bachawal, J. Willmann, Ultrasound molecular imaging: moving towards clinical translation, Eur. J. Radiol. 84 (9) (2015) 1685–1693. [57] H. Madjar, H. Prömpeler, C. Del Favero, A new Doppler signal enhancing agent for flow assessment in breast lesions, Eur. J. Ultrasound 12 (2000) 123–130. [58] M. Stuhrmann, R. Aronius, M. Schietzel, Tumor vascularity of breast lesions: potentials and limits of contrast-enhanced Doppler sonography, AJR Roentgenol. 175 (2000) 1585–1589. [59] B. Goldberg, C. Merritt, L. Needleman, et al., Contrast-enhanced ultrasound and mammography for breast cancer detection,”, J. Ultrasound Med. 20 (2001) S95. [60] C. Wan, J. Du, H. Fang, F. Li, L. Wang, Evaluation of breast lesions by contrast enhanced ultrasound: qualitative and quantitative analysis, Eur. J. Radiol. 81 (4) (2012). pp. e444-50. [61] Q. Hu, X. Wang, S. Zhu, et al., Meta-analysis of contrast-enhanced ultrasound for the differentiation of benign and malignant breast lesions, Acta Radiol. 56 (1) (2015) 25–33. [62] P. Shankar, P. Krishna, V. Newhouse, Advantages of subharmonic over second harmonic backscatter for contrast-to-tissue echo enhancement, Ultrasound Med. Biol. 24 (1998) 395–399. [63] F. Forsberg, C. Piccoli, D. Merton, Breast lesions: imaging with contrastenhanced subharmonic US—initial experience, Radiology 244 (2007) 718–726. [64] J. Dave, F. Forsberg, S. Fernandes, Static and dynamic cumulative maximum intensity display mode for subharmonic breast imaging: a comparative study with mammographic and conventional ultrasound techniques, J. Ultrasound Med. 29 (8) (2010) 1177–1185. [65] J. Eisenbrey, J. Dave, D. Merton, et al., Parametric imaging using subharmonic signals from ultrasound contrast agents in patients with breast lesions, J. Ultrasound Med. 30 (2) (2011) 85–92. [66] A. Sridharan, J. Eisenbrey, P. Machado, et al., Quantitative analysis of vascular heterogeneity in breast lesions using contrast-enhanced 3-D harmonic and subharmonic ultrasound imaging, IEEE Trans. Ultrason. Ferroelectr. Freq. Control 62 (3) (2015) 502–510. [67] T. Kim, A. Giuliano, G. Lyman, Lymphatic mapping and sentinel lymph node biopsy in early-stage breast carcinoma, Cancer 106 (1) (2006) 4–26. [68] B.B. Goldberg, D.A. Merton, J.B. Liu, et al., Contrast-enhanced ultrasound imaging of sentinel lymph nodes after peritumoral administration of Sonazoid in a melanoma tumor animal model, J. Ultrasound Med. 4 (30) (2011) 441– 453. [69] S. Rautiainen, M. Sudah, S. Joukainen, et al., Contrast-enhanced ultrasound guided axillary lymph node core biopsy: diagnostic accuracy in preoperative staging of invasive breast cancer, Eur. J. Radiol. 84 (11) (2015) 2130–2136. [70] K. Omoto, H. Matsunaga, N. Take, et al., Sentinel node detection method using contrast-enhanced ultrasonography with sonazoid in breast cancer: preliminary clinical study, Ultrasound Med. Biol. 35 (8) (2009) 1249–1256. [71] J. Eisenbrey, A. Sridharan, P. Machado, H. Zhao, Three-dimensional subharmonic ultrasound imaging in vitro and in vivo, Acad Radiol. 19 (6) (2012) 732–739.