February 25, 2008
Automated Breast Ultrasound — Increasing Sonography’s Reproducibility
By Kathy Hardy
Vol. 9 No. 4 P. 10
Combine the known benefit of ultrasound for breast cancer diagnosis with an automated image acquisition tool that can simplify the ultrasound process and help produce consistent results and you have what some physicians are calling the best thing since sliced imaging.
The technology at the forefront of this reaction is SomoVu, an automated breast ultrasound system developed by U-Systems Inc. This system integrates an automated transducer arm, rather than the standard handheld probe, with a display workstation. The acquisition process allows the technologist to select individual diagnostic planes that are captured in an automated scan. The display workstation provides complete 3D renderings of transverse, sagittal, and coronal views. The coronal view is most unique because such an angle is not possible with 2D ultrasound.
“It’s the icing on the cake,” says Dennis Yutani, MD, a radiologist at Memorial Hospital of Converse County in Douglas, Wyo. “Automated breast ultrasound takes what was a good imaging modality and makes it better.”
Establishing Its Role
Silicon Valley-based U-Systems introduced SomoVu in 2006 after 10 years of research and development, according to Bob Thompson, the company’s interim president. The automated ultrasound system is currently used in 30 imaging centers and hospitals worldwide, plus 10 additional clinical trial sites in the United States. SomoVu has been FDA approved for diagnostics but not for breast cancer screening.
“There’s no data that says this can replace other screening modalities,” Yutani says. “Automated breast ultrasound is an enhancement to the process and particularly helpful for women with dense breast tissue or women with breast implants. I would hesitate to say we should use this for screening until there is more data out.”
From a procedural standpoint, the patient experience with automatic ultrasound begins in a similar fashion as standard handheld ultrasound. The difference starts once the device is secured and engaged and involves how the breast tissue images are acquired. As Thompson describes it, the ultrasound data are acquired with a 14-centimeter transducer that is larger than the approximately 4-centimeter handheld transducer used with most ultrasound equipment.
“The larger transducer gives you a better spectral relationship,” Thompson says. “You can see more of the breast.”
Once the technologist positions the transducer over the breast, it automatically centers and then performs a sweeping breast scan. A preview scan is generated and automatically determines the patient’s breast tissue signature. That signature is necessary for setting the proper ultrasound imaging parameters for each patient, and the actual acquisition scan follows. Altogether, the process takes approximately 60 seconds.
During the scanning process, the technologist watches the acquisition to ensure proper breast coverage and tissue contact. When the scan is complete, the technologist reviews images on the monitor, confirms the nipple location, and sends the images to the display station for the radiologist to review. This scan captures 350 ultrasound images that are then rendered and viewed in 3D.
“The radiologist looks at the volume of data in its original planes in all views in three dimensions,” Thompson says. “The scan gives you a complete data set.”
“With the 3D rendering,” Yutani adds, “there is a large volume of data. It slices like salami and then the data is reconstructed. That’s the biggest part. You can see the different perspectives of what’s going on in the breast. It’s the same anatomy, but you’re seeing it better.”
It’s the better view that Jessica Guingrich, MD, first speaks of when discussing automated breast ultrasound. Guingrich is the medical director for the breast imaging section of the Susan G. Komen Breast Center in Peoria, Ill. This was the first site in the United States to use SomoVu and among the 10 initial sites that participated in a clinical trial that began in 2005 and continued until June 2007.
“The added benefit with automated breast ultrasound is the 3D look, particularly the coronal view,” she says. “That view is the most helpful. This is a view of the tissue from the skin to the chest wall.”
Guingrich, a medical advisor for U-Systems, reported results of the first phase of the two-year clinical trial at RSNA 2006. In that study, the automated breast ultrasound system performance was compared with conventional handheld ultrasound results in 165 women who had first undergone mammograms. Findings with mammography and automated ultrasound agreed with mammography and handheld ultrasound in 94% of the 177 breasts analyzed.
In an RSNA 2007 presentation, Guingrich discussed a trial in which the SomoVu device was evaluated to determine if the system could improve breast cancer diagnosis. In the study, women were first diagnosed using handheld ultrasound and then with automated ultrasound. In the end, Guingrich reported that the addition of 3D ultrasound imaging to the breast cancer diagnosis process could increase the chances of finding cancerous lesions.
While the ultrasound technology brings the 3D imaging to life, it’s the automation that adds consistency to the process. According to U-Systems vice president of marketing and clinical training Jeanine Rader, handheld ultrasound is dependent on the technologist manually manipulating the transducer over the patient’s breast. Exact comparisons between breasts, as well as comparisons with past or future ultrasound breast images, are more difficult with the handheld procedure.
“It’s more of an art and less science,” she says. “Our hope with this technology is to make it more of a science and less of an art.”
Automation creates standardized views based on how the technology of the larger transducer scans a patient’s breast, with imaging parameters for the breast scan determined by the breast size. The technologist can easily repeat this process for follow-up scans, allowing for reproducible image quality and consistency. Images can be saved and reviewed later for comparison purposes.
“Automation removes user variability,” Guingrich says. “With the handheld method, the view can be altered depending on the angle and how the transducer is held.”
Ultrasound technologists can also physically benefit from the scanning process automation. Repetitive stress symptoms in the hands and wrists are an issue among sonographers who have been performing handheld ultrasounds for many years. “Automation of the scanning process opens the door for sonographers who had left the field due to repetitive stress injuries,” says Andrea Roeder, a product manager with Siemens Medical Solutions Ultrasound Division, a SomoVu distributor since March 2007.
Yutani and Guingrich agree that this new technology doesn’t mean the end of handheld ultrasounds. “I still go back to the handheld ultrasound for another view if I see something after the automatic ultrasound is done,” Yutani says.
Guingrich adds that automatic ultrasound provides a good overview of the breast tissue, but the handheld can be used for targeting one specific area for further review.
“The handheld ultrasound provides a good evaluation for specificity, but when you’re looking for an overview, the automatic ultrasound provides a good view, particularly with dense breast tissue,” she says. “I’ll still use handheld for a correlation to compare with the automatic scan.”
New View for Radiologists
Thompson says that as with any imaging modality, there needs to be some up-front training for the technologists performing the scans, as well as for radiologists concerning how to read the results. For example, he says that with the automatic ultrasound system, the coronal view can take some adjustment by the radiologist to interpret.
“It also takes some adapting to view in 3D,” he says. “People think in 3D, but they view in 2D. There is a learning curve. It depends on their experience using 3D. It varies individual to individual.”
Guingrich agrees that physicians need to adjust to viewing images from three perspectives rather than one. “There are a lot of images there, with views obtained at different angles,” she says. “Once your eyes become accustomed to the slices, then the reviewing process goes pretty quickly. You have to be methodical. As you read more, you begin to apply your own pattern for viewing ultrasound slices. This is true with all modalities.”
Guingrich gives one tip for technologists preparing the patient for an automated exam. “There is a learning curve for the technologists in applying the gel to the patient’s breast prior to scanning,” she says. “You need no air pockets. The more uniform the gel application, the better the view.”
“From the technician’s standpoint, it’s pretty straightforward,” Yutani says. “For the radiologist, you’re looking at hundreds of slices. We’re still in the learning phase as far as knowing what the artifacts look like. To say you look at all views at 100% is generous. I focus on the initial view and the coronal view closely.”
Yutani is the sole radiologist for Memorial Hospital, a 25-bed critical care facility that has been labeled a “Rural Site of Excellence” for SomoVu because of the hospital’s work with U-Systems in integrating automated ultrasound with PACS. Yutani and a staff of 12 technologists began using the equipment in December 2007 and have adapted to its nuances, he says. He and his department have experience with leading-edge technologies; their imaging department is “fully digital,” and they have a 64-slice CT scanner and an in-house MRI scanner. This is considerable for a Rocky Mountain community of approximately 5,000 people that is the so-called jackalope capital of the world.
“We have a board and an administration that believes in our mission statement: ‘Providing advanced medicine and home town care,’” Yutani says.
Developing Automated Ultrasound Uses
Product manager Jackie Bailey, also with Siemens, has seen acceptance of the automated ultrasound technology from hospitals and imaging centers in the United States and internationally. “Siemens is committed to the development of different advancements in ultrasound technology,” Bailey says. “We saw how automated ultrasound could increase workflow in breast imaging and thought it was time to explore this and other solutions.”
Thompson says U-Systems will develop more products focused on breast imaging, as well as complete existing clinical trials. “We’re continuing to focus on the breast,” he says. “There’s a long way to go. We have a continued interest with the discovery of disease in breast tissue.”
“Ultrasound will never go away,” Guingrich adds. “Hopefully, there will be more acceptance of automated breast ultrasound and future integration with mammography. Hopefully, this will become more of a standard operating procedure than an exception.”
— Kathy Hardy is a freelance writer and editor based in Phoenixville, Pa.