Processing, technology improve breast imaging
Higher frequencies, better transducers, and new techniques portend an era of faster speeds and greater detail
By: BRIAN S. GARRA, M.D.

Nowhere in the past decade has the improvement in ultrasound image quality been more dramatic than in breast and small parts imaging. Much of this gain in tissue contrast and detail results from new digital signal processing that has made advanced applications both faster and more affordable.

But it is often forgotten that the breast was one of the first body parts to which ultrasound was applied. It was nearly 50 years ago that Wild and Reed used A-mode scanning to detect a breast tumor at 15 MHz. Since then, breast ultrasound has evolved from static B-scanners, through water-path immersion scanners, to real-time hand-held scanning, first at 5 MHz, then 7.5 MHz, and now at 10 MHz and above.

The steady improvement in image quality has spurred some practitioners to change their methods of scanning and interpretation to capitalize on the information that is now available. Despite the landmark FDA approval in 1995 that allowed the use of ultrasound in the diagnosis of solid breast masses, too many physicians and sonographers cling to obsolete methods, using ultrasound only to distinguish solid from cystic masses. For any modality to reach its diagnostic potential, the rules for conduct and interpretation of an imaging modality must change along with the technology.

Recent advances in breast ultrasound technology include:

- An increased number of transducer elements in ultrasound transducers, leading to higher lateral resolution.

- Higher center frequencies in the 10 to 15 MHz range, leading to higher lateral resolution.

- Broadband transducers and increased scanner bandwidth, resulting in higher axial resolution and allowing shifting the center frequency and frequency compounding to reduce noise.

- Increased sophistication of signal processing routines, leading to better images with lower noise and higher contrast.

- Improvements in Doppler sensitivity and power Doppler, making evaluation of the vascularity of breast masses relatively straightforward. When Doppler evaluation was first promoted in the early 1980s, continuous wave Doppler was used because of its higher sensitivity. Vessels around a breast mass were examined for high velocity flow that was thought to suggest malignancy. More recently, simply looking at the number of vessels in and around a mass with color Doppler imaging has accomplished the same purpose. The interpretation of Doppler information is still controversial, but most investigators and the ATL/FDA trial agree that increased vascularity suggests malignancy.

- The biggest single advance in recent years has been 1.5-dimensional scanner arrays, which provide dynamic focusing in the slice thickness plane, perpendicular to the normal scan plane. This feature dramatically improves lesion contrast close to the transducer, where most breast imaging occurs. Figure 1 shows the improvement in image quality achievable with a 1.5-D array, compared with a conventional one-dimensional array in a tissue mimicking phantom, and in a patient.

NEW SCAN TECHNIQUES

Along with improvements in image quality have come procedural advances that take clinical advantage of the better scanners. The most important advance is the widespread acceptance of ultrasound for diagnosis of solid masses and the detection of cancer. New techniques for scanning have also appeared, including radial scanning, compression, and palpation during scanning.

- Radial scanning is critical for the detection of intraluminal mammary duct lesions. Figure 2 illustrates a solid mass within a mildly dilated mammary duct. If not viewed along the long axis of the duct, the mass is difficult to detect, but seeing the mass within the duct is relatively simple when the transducer is aligned along the duct.

- Compression is important for characterization of solid masses, since cancers compress much less than benign lesions. At times, solid masses may be isoechoic relative to the surrounding tissue especially when surrounded by fatty tissue, and looking for an area that does not compress as much as the adjacent tissue will help to detect the lesion.

- Palpation during scanning allows for precisely localizing palpable abnormalities relative to the ultrasound image. Palpation enables the examiner not only to find subtle lesions but also to determine when normal structures such as fat lobules and thickened Cooper's ligaments are causing a palpable abnormality. With this information, the examiner can often reassure an anxious woman about a palpable "lump."

Along with new techniques for ultrasound scanning, new institutions and organizations have arisen to help women obtain the highest quality breast ultrasound scans. The ARDMS Registry in Breast Ultrasound will emphasize modern scan techniques and identify individuals qualified to perform breast ultrasound. The AIUM breast ultrasound accreditation program helps ensure that breast imaging centers not affiliated with general imaging centers are performing high- quality breast ultrasound. Further work remains to be done on training centers for moving new technologies into clinical practice

NEW TECHNOLOGIES

Some of the new technologies that have been developed recently or will become available in the near future are harmonic imaging, ultrasound contrast agents, elastography, sonoelastography, two-dimensional transducer arrays, and three-dimensional breast ultrasound.

- Harmonic imaging is a procedure in which the ultrasound machine scans images at twice the frequency transmitted. This technique potentially can suppress reverberation and other near-field noise but it may limit depth of penetration and result in loss of resolution, unless the newer broadband harmonic imaging techniques are used. Figure 3 shows a breast cyst with internal echoes and illustrates the decrease in internal echoes under harmonic imaging.

Harmonic imaging has been shown to reduce the number of possible complex cysts or solid masses seen at breast sonography and improve the examiner's confidence that a lesion is in fact truly cystic and benign. The procedure also shows potential to better define the boundaries of lesions-an important feature in distinguishing benign from malignant lesions.

- Ultrasound contrast agents usually employ encapsulated bubbles or solid particles in the 5 to 7-micron range, producing a marked increase in backscatter and making it easy to visualize flowing blood. They also produce moderate tissue enhancement usable for dynamic perfusion studies that look for changes in tissue enhancement over time. When some agents are exposed to a higher power ultrasound beam, the microbubbles break, releasing acoustic energy that can be detected using color or power Doppler. This phenomenon has been called stimulated acoustic emission and may be useful for detecting contrast agents in tissue when the gray-scale imaging does not clearly show the agent.

The disadvantages of contrast agents are their cost and the requirement for an intravenous injection. Also, with more sensitive Doppler instrumentation, blood flow enhancement may not be as important as it has been in the past. Figure 4 shows the increase in tissue enhancement that can be achieved using an ultrasound contrast agent in experimental tumors implanted in a liver. Whether such enhancement will occur in breast applications remains to be seen.

- Vibrational Doppler imaging, or sonoelastography, is an outgrowth of the well-known fremitus breast ultrasound technique (Figure 5). To perform a fremitus examination the patient is asked to hum a pitch while color or power Doppler is used to examine the breast. Softer portions of the breast vibrate more in response to the humming, while cancers and other firm masses vibrate less and thus become visible as areas of decreased color, even if they are isoechoic on the ordinary B scan. Sonoelastography or vibrational Doppler imaging is similar, but an external transducer separate from the ultrasound transducer applies the vibrations. In the vibrational Doppler imaging approach, the external transducer is vibrated at various frequencies and the amount of tissue vibration at each frequency is quantified, using a quantitative power Doppler algorithm built into the scanner. Vibrational Doppler imaging looks at the viscoeleastic properties of tissue.

Preliminary results in about 20 patients indicate that the amount of vibration occurring in a lesion depends on the frequency in a complex manner. Lesions may or may not vibrate more than the surrounding tissue at any one frequency. This fact probably accounts for the variable results obtained using fremitus clinically. Different patients hum at different pitches and lesions may or may not vibrate differently from the surrounding tissue at a given frequency. Also, benign lesions showed wide variations in the amount of vibration as a function of frequency. Cancers, on the other hand, tended to vibrate less and showed much less variation as the frequency of the applied vibration was changed.

Using these differences between benign and malignant masses, it is possible to correctly identify most benign lesions without mistaking some cancerous lesions as benign. Figure 6 shows the spectra of two lesions, one benign and one malignant, with the increased variability as a function of frequency for a benign lesion and the decreased variability and decreased overall signal intensity for the malignant lesion. Note that at certain frequencies the lesions vibrate with the same intensity and thus would be indistinguishable from one another.

- Elastography looks at only the elastic properties of tissues by applying a slight compression to the tissue and comparing an image obtained before compression and after compression. The data collected before and after compression are compared, using a cross-correlation technique to determine the amount of displacement each small portion of tissue undergoes in response to the compression applied by the ultrasound transducer. The compression is very small, usually only 0.2 to 0.6 mm. The rate of change of displacement of the breast tissue as a function of distance from the transducer causing the compression is called a strain image and constitutes the elastogram.

Preliminary work in elastography has demonstrated that benign lesions such as fiberadenoma and fibercystic nodules tend to be either invisible or barely visible on an elastogram. Figure 7 pictures a typical fibroadenoma. Cancers, being much harder than the surrounding breast tissue, stand out readily on an elastogram (Figure 8).

Three different patterns have been identified in elastograms of cancers: a well-defined, very hard (dark) mass or nodule; a moderately hard mass or nodule containing much harder (darker) foci within it; and a very dark or hard central core surrounded by a somewhat softer or less dark peripheral component. One area where elastography may be of great benefit is in distinguishing fibrosis of the breast from cancer. Fibrosis generally appears as echogenic regions with posterior acoustic shadowing-an appearance often seen in cancers. In elastography, however, fibrosis generally shows as a uniform, moderately hard region with no distinct foci of increased hardness, whereas a cancer usually stands out as a well-defined though irregular region of increased hardness.

Preliminary work in elastography has shown that elastography can correctly classify most benign and malignant masses. This can be done on an elastogram because malignant masses appear darker or harder than benign masses and they also appear larger in the transverse dimension than benign masses would appear. Malignant masses appear larger on elastograms than their corresponding sonogram because of the surrounding desmoplastic reaction, which produces a zone of increased hardness surrounding the actual cancer.

In fact, in a study of the elastographic appearances of cancers and benign nodules, no cancer measured smaller in transverse dimension than the corresponding lesion did on sonography. Benign masses, however, frequently measure either smaller or the same size as their sonographic image. Using the overall signal intensity (hardness) of a lesion and the difference in size between the lesion and its sonographic image, it is possible to correctly classify most solid masses as benign or malignant. In prior work, these two features produced an area under the ROC curve of 0.86 for distinguishing benign from malignant, and six of 10 benign lesions were correctly classified as benign without missing any cancers. This technique may obviate many biopsies of benign lesions and reduce the cost and discomfort of diagnosing breast cancer. Over the past few years, elastographic image quality has improved, along with the sharpness of margins and the resolution within the interior of hard lesions.

- Two-dimensional transducer arrays can now produce three-dimensional ultrasound images. Of course, 3-D image segments may be produced using other technology, but the technique is often cumbersome. An important advantage of 3-D ultrasound is that it will allow for more rapid and reproducible scans and may solve the problem of screening ultrasound. Screening ultrasound has great potential but screening the entire breast sonographically is labor-intensive and time-consuming for both sonographer and physician. A screening test should be simple, relatively cheap, and ideally should not require a physician's presence. Screening 3-D ultrasound by a technologist or sonographer would permit the radiologist or other physician to review the 3-D data set in multiple scan planes, including the radial planes.

With new technologies based on ultrasound on the verge of practical usefulness, the practice of breast ultrasound will change if sonographers and sonologists embrace the new technologies, and not avoid them. Integrating ultrasound breast imaging with new non-ultrasound technologies is important, and so are programs to train and accredit as many individuals and practices as possible.

DR. GARRA is vice chairman of radiology at the University of Vermont in Burlington.

Bibliography

Blohmer JU, Bollmann R, Heinrich G, et. al. Three-dimensional ultrasound study (3-D sonography) of the female breast. Geburtshilfe Frauenheilkd 1996;56:161-165.

Chaudhari MH; Forsberg F; Voodarla A, et. al. Breast tumor vascularity identified by contrast enhanced ultrasound and pathology: initial results. Ultrasonics 2000;38:105-109.

Fornage BD. Recent advances in breast sonography. JBR-BTR, 2000;83(2):75-80.

Garra BS, Cespedes EI, Ophir J, et al. Elastography of breast lesions: initial clinical results. Radiology 1997;202:79-86.

Huber S, Helbich T, Kettenbach J, et. al. Effects of a microbubble contrast agent on breast tumors: computer-assisted quantitative assessment with color Doppler US-early experience. Radiology 1998;208:485-489.

Wright IA, Pugh ND, Lyons K, et. al. Power Doppler in breast tumors: a comparison with conventional colour Doppler imaging. Eur J Ultrasound 1998;7(3):175-181.

Back to top