A team of researchers is studying the use of stereographic imaging technology and three-dimensional (3-D) displays to detect potential breast malignancies missed by traditional mammographies, opening the door to earlier detection and treatment as well as reducing the number of false-positive results and follow-up tests.
Stereo mammography provides radiologists with a three-dimensional view of the internal structure of the breast by taking two images from slightly different angles—much the way our two eyes create depth perception, or moviemakers create 3-D IMAX films. These mammography images are displayed on two Planar Systems high-resolution—2,500-by-2,000 pixel—liquid crystal display (LCD) monitors attached one on top of the other at a 110-degree angle, with a specially coated glass partition between them. The glass allows a radiologist wearing cross-polarized glasses to see the lower monitor (placed at eye level) while simultaneously viewing a reflection of the second monitor (placed slightly above eye level and angled downward).
"The brain is able to put together these images taken from different vantage points and figure out where things are in terms of depth," says David Getty, division scientist at Cambridge, Mass.–based BBN Technologies, Inc., which developed the stereo mammography technology used during the trial. BBN began studying different technologies that could be applied to mammography in 1992.
According to study co-author Dr. Carl D'Orsi, a radiology professor at the Emory University School of Medicine in Atlanta, the new procedure would make it easier to find tiny tumors obscured by other tissue in the breast. "You're looking for something at the limits of human visibility," he says. "In other areas of the human body, you don't have to look for something that small" when screening for cancer. As a result, overlapping or neighboring healthy tissue can resemble malignancies or, conversely, may hide small growths.
As of July 2007, 1,093 patients at elevated risk for developing breast cancer were enrolled in Emory's clinical trial and had received both standard and stereoscopic digital mammography exams. A total of 259 suspicious findings were identified by the two tests and the patients were referred for additional diagnostic testing; 109 of the flagged spots turned out to be actual lesions. Standard mammography missed 40 of the lesions that the stereoscopic exam found, whereas 24 slipped by the stereoscopic exam but were found by standard mammography. "It is possible that there were a few other lesions that existed and that were missed by both modalities," Getty says. Still, stereoscopic digital mammography reduced the amount of false-positive findings turned up by standard digital mammography by 49 percent.
But the images are not the only problem. New research shows that the ability—or lack of ability—to properly interpret mammograms also plays a role in their effectiveness. A report by Seattle-based Group Health Center for Health Studies, a nonprofit health care system, published this week in the Journal of the National Cancer Institute indicates that accuracy of readings depends on the experience and skill of radiologists interpreting them: Those who read diagnostic mammograms most accurately tended to be based at academic medical centers and / or have spent at least 20 percent of their time doing such assessments. Most mammograms in the U.S., however, are interpreted by general radiologists, who only spend a fraction of their time analyzing such x-rays.
That is not to say mammograms are easy to read. A radiologist or doctor often is hunting for early-stage cancer formations less than a half-centimeter (0.20 inch, or 0.50 centimeter) in diameter. Still, it is a skill in great demand. In 2004 (the most recent year for which numbers are available) 186,772 women and 1,815 men were diagnosed with breast cancer, according to the U.S. Department of Health and Human Services' Centers for Disease Control and Prevention.
Stereoscopic mammography systems could be built by simply adding a stereo display to existing digital mammography equipment, Getty says, noting that he and colleagues are meeting with mammography equipment manufacturers to gauge their interest and, also, are seeking funding for further technology development and clinical trials from the National Institutes of Health.
Stereo mammography holds even greater promise as mammography equipment manufacturers such as General Electric Co., based in Fairfield, Conn.; Siemens, AG, headquartered in Munich; and Bedford, Mass.–based Hologic, Inc., develop machines to perform breast digital tomosynthesis, which takes up to 20 images in an arc in front of each breast with each image separated by one or two degrees. Together, these images could be reconstructed and viewed through a stereo mammography system to create a 3-D image of the breast that can be examined from a number of different angles.
"Breast CT [computed tomography] scanning may be the ultimate, but that's a decade away," D'Orsi says. Getty is hoping that ultimately CT or magnetic resonance imaging (MRI) scans of internal organs—such as the prostate and lungs—will also be viewable this way.