Inside the box

Posted on 21. Jun, 2011 by in Academic Departments, Electrical and Computer Engineering, Features, Healthcare and Medicine, Issues, Magazine, Research

New method for breast cancer imaging

Every woman over the age of 40 receives the same initial screening for breast cancer: a mammogram. Yet no two women are identical and neither are their breast cancer risks, so a team of University of Wisconsin-Madison researchers is developing a system better tailored to women with a particularly high risk factor.

In 2000, a National Academy of Engineering publication identified breast cancer detection as a healthcare problem in need of an engineering solution. In the subsequent decade, Philip Dunham Reed Professor of Electrical and Computer Engineering Susan Hagness has emerged as a leader in the search for that solution. The system will offer three-dimensional capabilities similar to a magnetic resonance imaging (MRI) system along with the affordability and accessibility of traditional mammography.

Susan Hagness and Barry Van Veen

Susan Hagness and Barry Van Veen. Photo: David Nevala.

Hagness works closely with Electrical and Computer Engineering Professor Barry Van Veen, and their collaborators span a variety of fields. Electrical and Computer Engineering Assistant Professor Nader Behdad, Duane H. and Dorothy M. Bluemke Professor of Electrical and Computer Engineering John Booske, and Radiology Associate Professors Fred Kelcz and Gale Sisney currently are contributing to the project.

The density risk
Breast tissue is made up of fatty tissues, connective tissues and epithelial tissues, which line many of the body’s surfaces and cavities. Collectively, the connective and epithelial tissues are called fibroglandular tissue, and this tissue determines breast density. If a woman has a high percentage of fibroglandular tissue, her breasts are considered “dense.”

According to research published in the New England Journal of Medicine, high breast density can increase a woman’s risk for cancer to four to six times that of women with predominantly fatty tissue. It’s a stronger risk factor than early-onset menstruation or having no biological children. In fact, few other factors exceed dense breast tissue as a risk for cancer; a Radiology paper found those that do include a breast cancer gene mutation, age or prior breast cancer.

This strong risk is also fairly common. Around 50 percent of women in their 40s and 25 percent of women in their 70s have breast tissue that is at least 50 percent dense, according to the American Journal of Roentgenology. “All of these facts point to the importance of breast density evaluation in assessing a woman’s risk and having clinicians provide appropriate prevention protocols,” Hagness says.

Unfortunately, dense breast tissue makes it more difficult for doctors to accurately screen for cancer. Research in Annals of Internal Medicine found as many as two out of every five cancers in women who have high breast density go undetected. The problem is that mammography is a two-dimensional imaging technique. A mammogram machine takes a three-dimensional volume of tissue, passes X-rays through the tissue and creates a shadow gram. All of the tissue is projected onto that two-dimensional image.

Hagness compares it to trying to find a needle in a haystack. “It’s easier to find the needle if you can sift through the hay layer by layer; it’s much more difficult if you compress all of the hay into a thin pile and view it all at once,” she says.

However, Hagness and Van Veen don’t view their system as a replacement for traditional mammography. “Mammography is the gold standard that has saved countless lives, and we don’t see a need for an alternative for women who are served well by that technology,” Hagness says. “But there is a population that is currently underserved, and we’re interested in developing a safe, low-cost imaging modality that could be used for evaluating breast density and screening women who are at high risk.”

An affordable alternative for high-risk women
A three-dimensional image of dense breast tissue would allow doctors to sift through the entire tissue slice by slice. An MRI is an example of a system that can produce these kinds of images, but these scans cost around 10 times as much as mammograms.

Beyond the cost, the accessibility of MRIs is not ideal. Not all clinics or hospitals have MRI machines, and rural clinics are especially less likely to have one. In addition, a time-intensive MRI scan is a difficult experience for claustrophobic or obese patients, who make up a significant percentage of the population. Instead, Hagness and Van Veen are developing an imaging system that can produce three-dimensional images via
microwaves, a technology comparable in cost to a mammogram.

The clinical prototype will look like a box similar in shape and size to a vertical Kleenex tissue box, with tiny copper-colored antennas mounted on each side. Each antenna will transmit a low-power microwave signal,

A model of the new imaging system

A model of the new breast cancer imaging system. Electrical and Computer Engineering Assistant Professor Nader Behdad co-developed the antennas and sensor array with Hagness.

and all the other antennas will record the scattered signals from the breast. Algorithms will reconstruct those signals into a three-dimensional image of the breast tissue. Safety is a key component of the system. “This will transmit much less microwave power than a cell phone,” Hagness says. “It’s non-ionizing, so there is no health risk.”

Piloting the future of breast imaging
Achieving a clinical prototype has been an evolutionary process that has taken several years of research. From 2002 to 2007, Hagness led a large multi-institutional study to establish that breast tissue microwaves could convey important physiological information. The study involved measuring hundreds of freshly excised tissues from mastectomies, breast reductions and biopsies. The samples included healthy and cancerous tissues with a range of densities.

Hagness found that dense fibroglandular tissue does in fact have different electric properties from fatty tissues—a contrast that is necessary to evaluate breast density with microwaves. The work remains the definitive study on microwave properties of breast tissue, and the papers associated with the study have been cited hundreds of times. “We can distinguish dense tissue from fatty tissue, and that’s what you need to determine volumetric breast density,” she says.

Van Veen, who wasn’t directly involved in the tissue study, says the team’s work has evolved because of those findings. “Initially we had assumed tumors were these different objects that would scatter a lot more microwave signals than the healthy tissue,” he says. “But both tumors and healthy tissue scatter similar levels of energy back. That knowledge has changed the type of signal processing we do for this problem.”
The change means a shift from a radar-like approach to an imaging approach that looks for changes over time or changes due to contrast agents. With funding from the U.S. Department of Defense, Hagness and colleague John Booske are studying microbubbles and carbon nanotubes as possible contrast agents that will target tumors and make them more “visible” to the antennas.

Overall, the team is focused on moving the clinical prototype forward so the small-scale human trial can begin sometime in the next two years. This means developing and testing the algorithms that will actually generate images from the data gathered from the antennas. The team also is working on the actual sensor array that will be placed around the subject’s breast. “We’re beginning to make the transition from pure laboratory research toward clinical studies,” Hagness says.

The pilot will test around a dozen women. Participants will lie face down on an MRI support platform and place their breasts in a stabilizing structure that suspends the breast in the prototype “box.” The antennas will send a low dose of safe microwaves through the breast tissue, and the signals will be converted into a three-dimensional image. Participants will also undergo an MRI scan as a control test.

Applying engineering to healthcare
The field of electrical and computer engineering is becoming a more common home for healthcare research, says Booske, who chairs the UW-Madison Department of Electrical and Computer Engineering. “Currently, at least 13 of our faculty are pushing back the research frontiers in topics at the interface of ECE and biomedical or biological technology fields, and more get involved each year,” he says. “Being involved in this research on breast cancer detection and treatment has been one of the most exciting experiences of my career,” he adds. “My interests and expertise in the science of how electromagnetic waves interact with media has enabled me to contribute, and I’ve learned so much by working alongside the interdisciplinary team of experts brought together by Professors Hagness and Van Veen over the years on this initiative.”

Van Veen says the team’s overall breadth of expertise is what sets it apart from others working on similar research. “From state-of-the-art electromagnetics to sophisticated signal processing, we have a wide breadth that other research teams working on this problem don’t have,” he says. “We’re well positioned to hopefully ultimately solve the problem.”

Solving the problem is the fun part for Van Veen, who jokes about his broad range of signal-processing research experiences. “I like to say I’ve worked on problems from A to Z, though I haven’t quite gotten to Z,” he says. “I started my career applying signal processing to acoustics and have made it to W, wireless communications.”

When Hagness joined UW-Madison in the late 1990s, it wasn’t long before Van Veen began asking questions about her interest in breast cancer imaging, and the two began collaborating. “My expertise is essentially
in problems where multiple, different sensors simultaneously measure a physical phenomenon, such as electromagnetic scattering from breast tissue, ” he says.

In addition to her expertise in electro-magnetics, Hagness is a champion for increasing interest in engineering by showing students how the field can help solve global and societal challenges. She leads an introductory engineering course dedicated to the topic and works on major engineering outreach initiatives. “All engineering fields offer enormous opportunities to address a variety of challenges in the medical arena. The
reason why engineering offers this opportunity is because fundamentally, we’re problem- solvers,” Hagness says. “When a clinical need is identified, it’s a natural fit for engineers to try to address that need, and in electrical engineering, we have a lot of tools and skills to offer.”

“A significant part of the human body is electrical in nature. The nervous system, the brain, cell membranes—electrical engineers can fully understand that side of the human body,” she says.

Tags: , , , ,

No comments.

Leave a Reply

E-mail address