Ultrasonics 44 (2006) e217e220 www.elsevier.

com/locate/ultras

Critical issues in breast imaging by vibro-acoustography

Azra Alizad
a

a,*

, Dana H. Whaley b, James F. Greenleaf a, Mostafa Fatemi

Department of Physiology and Biomedical Engineering, Mayo Clinic College of Medicine, 200, 1st St. SW Rochester, MN 55905, United States b Department of Radiology, Mayo Clinic College of Medicine, 200, 1st St. SW Rochester, MN 55905, United States Available online 30 June 2006

Abstract Clinically, there are two important issues in breast imaging: detection of microcalcications and identication of mass lesions. X-ray mammography is the main imaging method used for detection of microcalcication, and ultrasound imaging is normally used for detection of mass lesions in breast. Both these methods have limitations that reduce their clinical usefulness. For this reasons, alternative breast imaging modalities are being sought. vibro-acoustography is an imaging modality that has emerged in recent years. This method is based on low-frequency harmonic vibrations induced in the object by the radiation force of ultrasound. This paper describes potential applications of vibro-acoustography for breast imaging and addresses the critical imaging issues such as detection of microcalcications and mass lesions in breast. Recently, we have developed a vibro-acoustography system for in vivo breast imaging and have tested it on a number of volunteers. Resulting images show soft tissue structures and calcications within breast with high contrast, high resolution, and no speckles. The results have been veried using X-ray mammography. The encouraging results from in vitro and in vivo experiments suggest that further development of vibro-acoustography technology may lead to a new clinical tool that can be used to detect microcalcications as well as mass lesions in breast. 2006 Elsevier B.V. All rights reserved.
Keywords: Vibro-acoustography; Breast imaging; Ultrasound; Mammography; Radiation force

1. Introduction Pulse-echo ultrasound (ultrasonography) and X-ray mammography are the common modalities used for breast imaging. Normally, ultrasonography is used to image soft tissue imaging and detect lesions in breast. ray mammography is the only imaging modality clinically used for detection of breast microcalcications. The sensitivity of mammography is greatly reduced in dense breasts. The accuracy of lm screen mammography is also inuenced by the experience of the radiologist, with experienced radiologists having the highest sensitivity in diagnosing breast cancer [1]. Furthermore, the ionizing nature of the X-ray mammography limits its frequent use. Ultrasonography application is hampered by the speckle patterns inherent to this imaging modality. Limitations in mammography

Corresponding author. Fax: +1 507 266 0361. E-mail address: aza@mayo.edu (A. Alizad).

and ultrasonography have prompted investigators to explore alternative breast imaging techniques. Especially, non-invasive imaging methods that can show both the soft tissue and microcalcications are of particular interest. Vibro-acoustography is a new imaging method based on the radiation force of ultrasound [2,3]. This method can be particularly useful for detecting hard inclusions in soft material. For example, vibro-acoustography has been used to image calcications in human arteries [46], microcalcications in breast tissue [7,8] and calcied arteries in breast [9]. A comparative study of vibro-acoustography with other radiation force methods for tissue elasticity imaging is presented in [10]. The transverse spatial resolution of vibro-acoustography is in the sub-millimeter range, making the technique suitable for high-resolution imaging [6 8,10,11]. The depth resolution, representing the slice thickness, is in the sub-centimeter range. Recently, we have developed a vibro-acoustography system for in vivo breast imaging and have tested it on a number of volunteers. Here, we describe applications of

0041-624X/$ - see front matter 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.ultras.2006.06.021

e218

A. Alizad et al. / Ultrasonics 44 (2006) e217e220

vibro-acoustography in breast imaging, and present preliminary in vivo breast vibro-acoustography results. 2. Method Vibro-acoustography is based on vibro-acoustic response of the object to a vibrating force [2,3,5,6,11]. Radiation force is generated by a change in the spatial distribution of the energy density of an incident acoustic eld. This can happen when an acoustic eld interacts with an object. The energy density of the impinging sound may change due to energy absorption, scattering, and reection. Thus, a radiation force is exerted on the object. The magnitude of this force depends on a number of parameters, including the scattering and absorption properties of the object. In a simple case of a plane wave reected from a planar object, the force is proportional to the power reection coecient of the object. In vibro-acoustography, we use two intersecting continuous wave (CW) focused ultrasound beams of dierent frequencies. The two ultrasound beams are focused and they are aligned to intersect at their focal region. At this intersection region, which is normally a small volume, the combined ultrasound eld energy density is sinusoidally modulated, and hence, the eld generates a highly localized oscillatory radiation force when interacting with parts of the object in this region. Thus, the resulting radiation stress is conned to a small region, which acts as an oscillating point force placed remotely inside the object. The general principle of vibro-acoustography is illustrated in Fig. 1. The confocal transducer produces two co-focused beams that generate an oscillating radiation force on the object at the dierence frequency. The hydrophone receives the resulting sound at the dierence frequency. Image of the object is formed from the output of the hydrophone as the beam scans the object. The object, transducer, and hydrophone are placed in a water tank. Amplitude and distribution of object motion is a function of its mechanical parameter such as the mass density, elasticity, and viscosity, as well as the boundary conditions, such as coupling to and the loading eects of the surrounding medium. The vibration motion results in a secondary acoustic eld (acoustic emission) that propagates in the object. The acoustic emission which is at Df frequency is

detected by an audio hydrophone. As the ultrasound beam is scanned across the object, the hydrophone signal is recorded and its amplitude is mapped into an image. A vibro-acoustography image depicts two types of information about the object: (1) ultrasonic properties of the object, such as the scattering and power absorption characteristics; (2) the dynamic characteristics of the object at frequency Df, which also relates to the boundary conditions and coupling to the surrounding medium [3]. The former properties are those that are also present in conventional ultrasound imaging. The latter properties, which are related to object stiness, can be described in terms of object mechanical impedance at Df. Such information is not available from conventional ultrasound. Another characteristic of vibro-acoustography relates to image speckle. Speckle is the snowy pattern seen in conventional ultrasound images. Speckles result from random interference of the scattered ultrasound eld. Speckles reduces the contrasts of ultrasound images and often limits one to see small structures, such as breast microcalcications in tissue. Vibro-acoustography on the other hand uses the acoustic emission signal, which is at a low frequency. The image in this modality is practically speckle free, resulting in high contrast images that allow small structures to be visible. This feature makes vibro-acoustography suitable for detection of breast microcalcications. In vitro breast tissue imaging by vibro-acoustography: Performance of vibro-acoustography in detection of breast microcalcications and breast arterial calcications has been studied through a series of in vitro experiments conducted on human breast tissues samples in a water tank [79]. 2.1. In vivo breast vibro-acoustography We recently developed a vibro-acoustography system for in vivo breast imaging [12]. This system is integrated in a clinical stereotactic mammography machine (Fischer Imaging Inc., MammotestTM system). The combined system is designed in such a way that it enables us to produce matching (from the same view angle) vibro-acoustography and mammography images of human breast. System parameters are: transducer frequency = 3 MHz, resolution 0.7 mm, scanning increments = 0.2 mm, ultrasound inten-

Hydrophone

Filter Detector

~ ~

Object

Confocal Transducer

1+
Confocal Transducers Scanning Motion of Beam

Image

Fig. 1. Vibro-acoustography system setup. Modied with permission from [12].

A. Alizad et al. / Ultrasonics 44 (2006) e217e220

e219

Ultrasound transducer X-ray tube

Breast

Patient bed Hydrophone X-ray detector

VA water tank

Slide

Compression panel
Fig. 4. In vivo vibro-acoustography image from the breast of a volunteer. (Left) vibro-acoustography scan at 2.5 cm from the skin. (Right) another vibro-acoustography scan at 3 cm depth. Reproduced with permission from [12].

Fig. 2. Combined vibro-acoustography-mammography system. Reproduced with permission from [12].

sity at the focal point = 700 mW/cm2 in compliance with the FDA recommendation for in vivo ultrasound. Fig. 2 shows combined mammography system. Patient lies in prone position on the examination bed with a breast hanging down through the hole. The breast is sandwiched between the back panel (X-ray detector), and a sliding compression panel that keeps the breast slightly compressed and xed for mammography and/or vibro-acoustography scanning. The transducer is moved away during mammography. Acoustic gel is applied to ensure proper acoustic coupling. 3. Results Vibro-acoustography images of a breast tissue sample with calcied artery is shown in Figs. 3. The image of calcied artery of breast is clearly detected as seen in X-ray of this sample. Fig. 4 demonstrates the in vivo vibro-acoustography image from the breast of a volunteer. These images cover a 5 5 cm area taken at the depth of 2.5 cm (left image) and 3 cm (right image) from the skin. The calcication (diameter approximately 1 mm) is seen in the left image as a bright spot in the top-left quadrant of the image on the left. The presence of this calcication was proven by mammography. Tissue structure is visible especially in the right image with remarkable contrast. The background in the left image is darker because the image brightness is adjusted to show the calcication which happens to be much brighter than the soft tissue. These images was acquired at frequency Df = 50 kHz. The scan time was about 7 minutes. This preliminary result demonstrates

Fig. 5. (a) Mammogram. (b) vibro-acoustography image acquired at Df = 50 kHz with the beam focused at the depth of 4 cm from the skin.

one can produce high contrast in vivo images at ultrasound intensities within the FDA guideline. These results also demonstrate that vibro-acoustography has enough resolution and contrast to show both microcalcications and the soft tissue. Fig. 5 demonstrates mammography and vibro-acoustography of a patient volunteer with a broadenoma in her left breast. The mammogram (a) showing a 5 5 cm area of the coronal view of breast of a volunteer. The breast includes a large calcication (bright spot) inside a broadenoma region. The broadenoma, which is barely visible around the calcication, is marked by a radiologist using the arrows. Vibro-acoustography image (b) acquired at Df = 50 kHz with the beam focused at the depth of 4 cm from the skin. The broadenoma containing the calcication is seen as a dark region. The arrows are imported from (a) to the corresponding region on (b), verifying that the dark region matches the radiologist reading of the broadenma in the mammogram. 4. Discussion An ideal breast imaging device must be able to image both calcications and soft tissue. It must also oer enough resolution and sensitivity for detection of microcalcications. The experimental in vitro and in vivo studies have demonstrated that vibro-acoustography has such capabilities. The spatial resolution can be improved by using transducer with higher center frequency. However, one must take into account the increase in tissue attenuation. The

Fig. 3. A breast tissue experiment: A=X-ray and B=vibro-acoustography of the breast tissue.

e220

A. Alizad et al. / Ultrasonics 44 (2006) e217e220

scanning speed may be improved by using array transducers to steer the beam electronically. Further investigation is needed to fully explore the potentials of vibro-acoustography for in vivo breast imaging. A number of considerations must be taken into account before implementing vibro-acoustography for clinical applications. For example, the coupling between the transducer and the breast must be suitable for clinical practice. Another consideration is the scanning time. This time may be too long for routine clinical applications. The scanning time must be short enough to avoid excess patient discomfort during imaging. A clinical vibro-acoustography system may be implemented based on contact array transducers. That is, instead of using a two-element confocal transducer used in the present study, an ultrasound transducer comprising of two two-dimensional arrays may be employed to produce the two intersecting beams needed for vibro-acoustography. The two beams from the arrays can be focused at a common focal point and steered rapidly across a given plane within the breast. These methods are currently being studied [13]. 5. Conclusions Vibro-acoustography has potential to provide newer information in breast imaging. Potential applications in vibro-acoustography breast imaging include microcalcications, calcied arteries, and detection of mass lesions. Future directions will include electronic beam forming with array transducers and tissue characterization. Acknowledgements The authors are grateful to the following individuals for their valuable work during the course of this study: Thomas M. Kinter for software support, Randall R. Kinnick for laboratory support and scanning tissues, Joyce Rahn for her help in scanning the patients, and Elaine C. Quarve for secretarial assistance. Supported by NIH

Grant EB-00535 and Grant BCTR0504550 from the Susan G. Komen Breast Cancer Foundation. Disclosure: Parts of the techniques used here are patented by MF and JFG. References
[1] L. Liberman, A.F. Abramson, F.B. Squires, et al., The breast imaging reporting and data system-positive predictive value of mammographic features and nal assessment categories, Am. J. Roentgenol. 171 (1) (1998) 3540. [2] M. Fatemi, J.F. Greenleaf, Ultrasound stimulated vibro-acoustic spectroscope, Science 280 (1998) 8285. [3] M. Fatemi, J.F. Greenleaf, Vibro-acoustography: An imaging modality based on ultrasound stimulated acoustic emission, Proc. Natl. Acad. Sci. USA 96 (1999) 66036608. [4] J.F. Greenleaf, R.L. Ehman, M. Fatemi, R. Muthupillai, Imaging elastic properties of tissue, in: F.A. Duck, A.C. Baker, H.C. Starrit (Eds.), Ultrasound in Medicine (Medical Science Series), Institute of Physics Publishing., Bristol, England, 1998, pp. 263277. [5] M. Fatemi, J.F. Greenleaf, Probing the dynamics of tissue at low frequencies with the radiation force of ultrasound, Phys. Med. Biol. 45 (2000) 14491464. [6] M. Fatemi, A. Manduca, J.F. Greenleaf, Imaging elastic properties of biological tissues by low-frequency harmonic vibration, Proc. of IEEE 91 (10) (2003) 15031517. [7] M. Fatemi, L.E. Wold, A. Alizad, et al., Vibro-acoustic tissue mammography, IEEE Trans. Med. Imag. 21 (1) (2002) 18. [8] A. Alizad, M. Fatemi, L.E. Wold, et al., Performance of vibroacoustography in detecting of microcalcications in excised human breast tissue: a study on 74 breast tissue samples, IEEE Trans. Med. Imag. 23 (3) (2004) 307312. [9] A. Alizad, M. Fatemi, D.H. Whaley, et al., Application of vibroacoustography for detection of calcied arteries in breast tissues., J. Ultrasound in Medicine 23 (2004) 267273. [10] M. Fatemi, J.F. Greenleaf, Imaging the viscoelastic properties of tissue, in: M. Fink, J.-P. Montagner, A. Tourin (Eds.), Topics in Applied Physic, 84, Springer Verlog, Heidelberg Berlin, 2002, pp. 257275. [11] M. Fatemi, J.F. Greenleaf, Imaging and evaluating the elastic properties of biological tissues, BMUS Bull. 8 (4) (2000), pp. 16 & 18. [12] A. Alizad, D.H. Whaley, J.F. Greenleaf, M. Fatemi, Potential applications of vibro-acoustography in breast imaging, Technology in Cancer Research and Treatment 4 (2) (2004) 151158. [13] G.T. Silva, J.F. Greenleaf, M. Fatemi, Linear array Beamforming for vibro-acoustography: a numerical simulation study, Ultrason. Imaging. 25 (2004) 117.