Ultrasound's Complementary Role in Breast Imaging

Breast ultrasound is making its mark - namely in characterizing palpable and non-palpable abnormalites found via mammography, determining whether a mass is fluid-filled or solid, guiding biopsies and interventions and examining problems in breast implants. The technology also is breaking new ground in radiation therapy planning, and a study is ongoing as to breast ultrasound's role in boosting breast cancer detection rates.

Current indications for breast ultrasound studies have not changed significantly over the past several years, but advances in equipment technology have enhanced image clarity and improved both diagnostic capability and facilitated interventional procedures. Ultrasound has proven a valuable adjunct to mammography in the diagnosis of breast cancer, particularly for women at high risk of developing the disease who also exhibit dense breast tissue.


Ellen Mendelson, MD, professor in the department of radiology and chief of the breast imaging section at the Feinberg School of Medicine of Northwestern University in Chicago, explains that the practice guidelines of the American College of Radiology dictate specific indications for the use of ultrasound in breast imaging. This imaging modality has proven valuable in the identification and characterization of palpable and non-palpable abnormalities and further evaluation of clinical and mammographic findings. Being able to detect whether a lesion is fluid filled or solid offers the clinician direction about next steps to take in the diagnostic process.

Ultrasound guidance of biopsies and other interventional procedures has proven to be an accurate method of performing these techniques.

William R. Poller, MD, FACR, associate director of the division of breast imaging in the Breast Care Center of Allegheny General Hospital in Pittsburgh, says, "Anytime we see anything on breast ultrasound, we may use a 25 gauge needle to aspirate contents up to an 8 gauge needle for biopsy."  He describes the necessity of being able to visualize the needle and the lesion with a system that is user friendly to facilitate accurate, fast sampling of the area of interest, followed by marking its location for future reference.

Poller uses the Philips Medical Systems iU22 system with a 5 to 12 MHz transducer.  He explains that higher frequencies produce better penetration, with greater detail. He uses power Doppler to examine vascularity of the lesion, and spatial compounding, called SonoCT, to enhance images.

Women with silicon breast implants may develop problems that require imaging visualization. While MRI is helpful in some circumstances, determining if a rupture has occurred outside the capsule is best imaged using ultrasound. Detecting whether a palpable mass is related to a wrinkle or bulge in the implant or an abnormality of the breast tissue is enabled through the use of this imaging technique as well.

A relatively new use of ultrasound involves treatment planning in preparation for radiation therapy. Mendelson explains that when breast conservation surgery has been employed, a fluid-filled area remains in the breast where the tumor was removed. The standard method of radiation therapy is to provide radiation to the entire breast, with a boost to the lumpectomy site. Ultrasound provides guidance about the correct location to deliver a boost of additional radiation for maximum effectiveness.

The last area where ultrasound becomes the imaging modality of choice involves the initial breast imaging technique for young patients under 30, lactating and pregnant patients. This reduces the need for x-ray exposure in this population of patients.


This spring, the American College of Radiology Imaging Network (ACRIN) in collaboration with the Avon Foundation, launched a multicenter clinical trial designed to evaluate the role of ultrasound as a supplemental screening tool for women with dense breasts as well as increased of developing breast cancer. ACRIN 6666 was first proposed by the principal investigator, Wendie Berg, MD, PhD, a breast imaging consultant at American Radiology Greenspring in Maryland, and was funded in part by the National Cancer Institute with half of the budget being provided by a grant from the Avon Foundation. Mendelson is co-director of this study. (See www.acrin.org/current_protocols.html#A6666)

Berg explains that the double-blind study protocol involves enrolling 2,808 women from 16 different imaging centers to receive an annual mammogram as well as independently performed and interpreted bilateral whole breast ultrasound at 0, 12 and 24 months. Besides the goal of evaluating the contribution of supplemental ultrasound, the study team will analyze the time and resources required to perform screening ultrasound.

Previous ultrasound studies accomplished in single center studies detected cancer in between 3 and 3.5 women per thousand  (a similar rate to mammography). This represents a doubling of the rate of cancer detection when ultrasound is added to mammography, and Berg notes that most of those studies did not look specifically at women with fatty breasts.

"The cancers that are seen only on ultrasound are on average one centimeter or smaller cancers," says Berg. "The vast majority are invasive, which is what you would expect. Mammography does well with calcifications, and most DCIS [ductal carcinoma in situ] has calcifications. DCIS is not seen well on ultrasound."

One of the other considerations included in the design of this study involved the expertise of the person who would be reading the study. Principal investigators designed a phantom with several tiny lesions in it and other educational materials. Clinicians involved in the study were required to scan the phantom and locate a certain percentage of the lesions and characterize them properly. In addition, they are required to review imaging studies from patients with confirmed lesions and correctly identify the lesions.

Besides meeting the educational requirement in order to participate in the study, centers were required to have ultrasound equipment with 5 cm broad bandwidth transducers that are capable of providing spatial compounding because that provides information such as the margin of lesions, and excellent detail. Berg says that all of the major manufacturers offer machines that meet the study requirements. She happens to use the Philips Medical Systems ATL HDI 5000, but says that some centers are using the GE Healthcare (Waukesha, Wis.) Logiq 900 unit, and others use the Siemens Medical Solutions (Malvern, Pa.) Acuson Sequoia, while still others are using the Toshiba America Medical Systems (Tustin, Calif) Aplio.

"In this trial, the mammogram and ultrasound images will be read separately," explains Berg. "Once we've recorded the results, one of the investigators reviews both studies together." This methodology is designed to eliminate any bias from knowledge of the results of one study when reading the other.

While ultrasound has performed well in the study, none of the experts anticipates replacing mammography as the gold standard of breast cancer imaging. Rather, by combining mammography with ultrasound for difficult-to-image women, the investigators hope to develop additional techniques to improve the detection rate.


John Lewin, MD, breast imager at Diversified Radiology of Colorado in Denver, describes the benefits of multiple beam angle transducers that allow a mass to be visualized more accurately. Some imagers employ power Doppler to determine vascularity of a mass, which provides another clue as to whether the lesion in question is benign or malignant. While not definitive, he notes color Doppler or Power Doppler adds one more piece of information.

The multiple beam angles are built into the transducer, which transmits sound beams at multiple angles, thus eliminating shadowing. Lewin says Philips calls the capability SonoCT. He has used the Siemens Medical Solutions Antares system, too, and finds advantages in its user-friendly interface and small footprint.

Thomas Stavros, MD, FACR, radiologist at Radiology Imaging Associates in Inglewood, Colo., is using the Toshiba Aplio system for his routine ultrasound breast studies as well as to participate in the ACRIN clinical trial.

"Both harmonics and spatial compounding are ways of improving signal-to-noise ratio," says Stavros. "To get better lateral resolution and better axial resolution, we push the frequencies, bandwidth and dynamic range as far as we possibly can." However, he notes that if he moves slightly beyond equipment limits, it can introduce artifacts into the image. But by using harmonics and spatial compounding, they can remove some of the artifacts.

Mendelson notes that broad bandwidth transducers provide lower frequencies capable of penetrating deeper tissues about 5 cm from the surface, as well as very high resolution in the near field. "These are phenomenal instruments, and the software for these systems is more sophisticated and beneficial than in the past."  She notes that all of the major manufacturers offer these features on their high-end systems.


Two new approaches to breast ultrasound can be considered works-in-progress.

Kevin M. Kelly, MD, director of breast imaging at Huntington Memorial Hospital in Pasadena, Calif., has been working to develop a semi-automated ultrasound scanner called SonoCine. The technologist is responsible for providing correct skin pressure and angle of inclination for the transducer, while the SonoCine system provides uniform speed and position. The transducer moves across the skin at 0.8 mm per second, with images collected at 800-micron intervals.

Using a Siemens Acuson Sequoia ultrasound scanner attached to a special frame, bed and carrier, the patient wears a camisole soaked with gel to hold a special gel pad in place over the nipple to reduce artifact as the scanner passes over that region. The camisole, made of a slightly elastic material, reduces breast motion as the transducer passes across the chest.  This system is designed to create reproducible images of bilateral breast scans accomplished at a uniform speed.

At this time, the system is entering the PMA (pre market approval) phase with the FDA. Kelly reports that they have begun to enroll patients with two or three centers presently involved, and an anticipated 16-center participation when the study is fully deployed. They continue to seek new centers to be included, and plan to enroll a total of 16,000 patients during this protocol. (See www.sonocine.com)

Siemens Medical Solutions Ultrasound Division, and U-Systems, Inc. (San Jose, Calif.), a developer of full field breast ultrasound (FFBU), have entered into a collaboration to integrate the SONOLINE Antares system into the U-Systems' FFBU unit.

The FFBU system was developed to provide whole-breast images presented in a standardized format, to enable high-volume, cost-effective ultrasound imaging on women with dense breasts. This ultrasound approach is intended to complement mammographic scans.

U-Systems offers proprietary hardware and software technology in their system that includes separate scanning and viewing stations. With the scan station, the breast is held in light compression (similar to mammography), while a high frequency 768-element transducer acquires 400 to 800 sagittal ultrasound images in a single sweep during the 60-second scan. Mammography technologists have been educated to perform these scans.

Once the images are captured, the viewing station provides one-to-one correlation between mammographic and ultrasound images from the two patient scans. This feature is designed to improve identification of mammographic abnormalities on the ultrasound scan. (See www.u-sys.com)


Although mammography remains the best screening test and is considered the gold standard for diagnosis of breast cancer, it is not perfect, especially in specific high-risk population of women. Ultrasound is capable of finding malignancies that mammography misses. Used in tandem, these two modalities show promise of detecting more cancers than either of them alone.