A novel breast cancer imaging technology that uses microwaves instead of X-rays to detect breast tumors is being developed by researchers at Northwestern University, the University of Wisconsin-Madison and Interstitial, Inc., an Evanston-based startup company.
The new microwave technology is based on a fundamentally different principle than X-ray mammography, which shadows potential tumors by passing high-energy ionizing radiation all the way through the highly compressed breast to expose film on the other side. In contrast, the microwave approach uses a miniature antenna contacting the skin surface to bounce very low energy, non-ionizing microwave pulses off potential tumors. No breast compression is required.
"We are essentially designing a breast tumor radar," said Allen Taflove, professor of electrical and computer engineering at Northwestern's Robert R. McCormick School of Engineering and Applied Science. Taflove directed extensive supercomputer modeling of the microwave sensor and aided in the development of the technology. "We believe that this technology has the potential to complement X-ray mammography by detecting tumors that are invisible to X-rays, while avoiding X-ray exposure and potentially uncomfortable breast compression."
Scientists have been searching for screening and diagnostic methods that could improve on X-ray mammography. According to the National Cancer Institute, X-ray mammograms miss between 10 percent and 30 percent of malignant tumors. Many of the suspicious masses they do detect turn out to be benign -- after costly and invasive biopsies. Ultrasound, MRI and new optical breast imaging techniques are generally not foreseen as front-line screening methods for healthy women.
Highly sensitive microwave detection may help to reduce breast cancer mortality by warning of malignant tumors at a very early stage. The anticipated low cost and painless application could make it useful for routine screening. Detailed computer simulations and tests of a laboratory-demonstration sensor suggest that the microwave technology may also help to discriminate malignant from benign tumors, thereby reducing the number of unnecessary surgical biopsies.
The microwave sensor works by detecting differences in water content between a malignant tumor and normal breast tissue. The device is the brainchild of Jack Bridges, chairman of Interstitial and an electrical engineer with 50 years of experience in industry and more than 60 patents to his credit. Bridges knew that microwaves interact with human tissues primarily according to water content. Since malignant tumors have a much higher water content than normal breast tissues, he conjectured that microwaves could provide the basis for a highly sensitive detection system, especially if an antenna could be constructed that exploits the "whispering gallery" phenomenon familiar to visitors of science museums. Bridges envisioned that a tumor's microwave echo could be tuned in back at the source of the radar pulse, while the noisy clutter from the complex surrounding normal tissues would be rejected.
"Once we had these key concepts, the technology began to develop very rapidly," said Susan Hagness, a former graduate student in Taflove's laboratory who received her Ph.D. in 1998. Hagness used Cray Research, Inc. supercomputers to model the microwave detection technology. "The use of supercomputers in designing the microwave sensor saved several years of development time," Bridges said.
"On the Cray supercomputers, we were able to simulate embedding a tumor within randomly oriented veins and mammary ducts and lobes in the breast," said Hagness, who is now an assistant professor of electrical and computer engineering at the University of Wisconsin-Madison and is continuing to work on the technique's development. "We found that the ability of the microwave sensor to detect small tumors was only slightly affected by natural variations in the properties of the breast tissues."
A complete microwave tumor-detection system will consist of a sensor and a computer processor. The microwave sensor will be a small flat panel containing an array of miniature antennas placed on the surface of the breast. "We can focus the microwave signal within the breast, and the entire breast can be electronically scanned under software control to create a three-dimensional microwave image," Bridges said.
A laboratory prototype of the microwave sensor was built almost entirely with off-the-shelf components. The device was successfully tested on a "phantom" -- a simulated breast consisting of materials that share key microwave properties of normal breast tissues. A tiny simulated tumor was embedded within the breast phantom and was located using the microwave sensor. Phantoms are well-established proving grounds for new techniques and are used for routine calibration of existing mammography equipment. The microwave sensor has not yet been tested on humans.
The breast-phantom tests showed that the laboratory prototype sensor could easily detect one-quarter-inch simulated tumors embedded about 1.5 inches deep. With modest equipment improvements, even smaller tumors at least two inches deep could be detected. These results were consistent with those developed by the supercomputer simulations. The results also suggest that microwaves can detect some early-stage tumors that are invisible to X-ray mammography. X-rays rely on the presence of tiny calcium deposits within tumors to sense tumors too small to cast a shadow; however, not all early-stage tumors contain such calcium deposits. On the other hand, microwaves do not require calcium deposits for detection. In the case of microwaves, a tumor's response is generated by its entire cross section.
"A conventional X-ray mammogram is literally a shadowgram," Taflove explained. "As X-rays pass through the compressed breast, shadows appear on the X-ray film. The shadows represent the projection of a complex three-dimensional array of tissues onto a two-dimensional film surface. This makes X-ray results difficult to interpret."
Yet despite X-ray mammography's drawbacks, it is proven to save lives. The American Cancer Society recommends yearly screening mammograms for women over 40. The ACS projects breast cancer to strike 178,800 women in the U.S. this year and cause 43,500 deaths. October is Breast Cancer Awareness Month.
It is vital that women not put off starting or continuing their regular screening mammograms while new technologies are in long-term development. Taflove, Bridges and Hagness all stressed this point.
According to the U.S. Food and Drug Administration, of the 80 million women in the U.S. who are recommended for screening, only 23 million undergo an annual mammogram. "Noncompliance may be due in part to women's concerns about the small risks of X-ray exposure and the discomfort of the breast compression required by mammography," Hagness said.
Microwave imaging would eliminate both concerns. "We expect the microwave sensor to use only a few milliwatts of power -- less than one one-hundredth of the power of a typical cellular phone," Bridges said. "Further, the duration of microwave exposure during the breast scan will be shorter than many phone calls and much, much less frequent."
The discomfort of X-ray mammography will also be eliminated, Hagness said. "Instead of compressing the breast between two plates, the microwave antenna array will gently rest flat against the breast much as a book would if you were lying in bed."
Data from microwave imaging will be inherently digital, with no film to develop or scan. "After signal processing, the operator will see a three-dimensional image of the breast that can be rotated," Taflove said. "It will be captured digitally on a hard disk and could be sent to a radiologist over the Internet."
Interstitial was awarded two U.S. patents this year on the new breast tumor radar technology. Several additional patents are pending.
The computational feasibility study was funded by a Phase-I Small Business Innovation Research grant from the National Institutes of Health to Interstitial with a subcontract to Northwestern. Prototype testing to date has been funded by Interstitial.
Results from the supercomputer simulations of the microwave tumor detection system were presented by Hagness in October 1997 in Chicago at the Institute of Electrical and Electronics Engineers International Conference of the Engineering in Medicine and Biology Society, and in June 1998 in Atlanta at the IEEE Antennas and Propagation Society International Symposium. Initial results will also appear in refereed papers in December 1998 in IEEE Transactions on Biomedical Engineering, and in early 1999 in IEEE Transactions on Antennas and Propagation.
Interstitial is a tenant company of Northwestern's Basic Industry Research Laboratory, part of the Northwestern University/Evanston Research Park. The park is a research environment that seeks to accelerate technology transfer from the laboratory to the marketplace by combining the resources of a major university, a progressive community, and private industry. The park has a large concentration of entrepreneurial companies in such areas as software development, information-based technologies, medical services, and biotechnology. The companies enjoy ready access to Northwestern faculty and graduate students and to laboratory and library facilities. Northwestern has been a leader in creating interdisciplinary research centers for startup businesses since 1982.
In 1989 Hagness was in the first entering class of the McCormick School's Honors Program in Undergraduate Research, a combined B.S./Ph.D. program. She received her B.S. in electrical engineering in 1993, graduating at the top of her McCormick class. She also received national honors, including being named the outstanding junior-year electrical engineering student in the U.S. for 1992.
Other Northwestern members of the technical team include Milica Popovic, a graduate student in Taflove's group, and Alan Sahakian, associate professor of electrical and computer engineering and biomedical engineering. Interstitial personnel include Marvin Frazier, Robert Moulton, Samuel Shelfo and Raymond Zalewski. Consultants include Dr. John Aarsvold of the Department of Nuclear Medicine at the Atlanta Veterans Administration Medical Center and Dr. Melvin Griem, professor emeritus of radiation and cellular oncology at the University of Chicago.