OHSU Scientists Develop MRI Approach to Improve Breast Cancer Detection
02/23/05 Portland, Ore.
Continued studies may result in newer, more effective breast cancer detection methods
Researchers at Oregon Health & Science University's Advanced Imaging Research Center (AIRC) are developing a new imaging method that may provide a clearer diagnosis of breast cancer. The research is published in the latest issue of the journal Magnetic Resonance in Medicine. AIRC Director Charles Springer, Ph.D., is senior author, and AIRC Manager, Xin Li, Ph.D., is first author of the new paper, along with William Rooney, Ph.D., AIRC faculty. Professor Springer also holds appointments in OHSU's Cancer Institute and Department of Biomedical Engineering.
"This technique involves a new method for interpreting information gathered through MRI," explained Springer. "The technique involves recognizing that certain properties of MRI signals can change during the examination, much like the changing of a camera's shutter speed. On a camera, a fast shutter speed can make a speeding car look as if it is standing still. A slower shutter speed may result in a photo showing the car blurring past the camera. This principle, when correctly applied to MRI imaging, can provide more accurate information. In the case of MRI, the blurring is not of the actual image, but of the time courses of the MRI signals."
Magnetic resonance imaging technology combines the use of powerful magnets and radio wave pulses. The magnet influences the magnetization of the body's water molecules. The radio signals that are received from this can be converted into a visual representation. The top image is from a patient who had a false positive mammogram reading. In the OHSU shutter-speed map, no lesion is visible. The second image (middle) is from a patient who had a positive mammogram reading. The lesion is very clearly seen. In this case, the biopsy proved to be positive and the shutter-speed model was again correct. The third image (bottom) is a close-up image of the malignant tumor. The three red "hot spots" are very tiny - two to three mm in diameter.
The shutter speed concept allows researchers to adjust the mathematics of the computer program analyzing the signals to account for the movement of water molecules in and out of cellular compartments in diseased and healthy tissue. When the MR shutter speed increases, this movement appears to slow. In the case of tumors, using shutter speed analysis not only more clearly indicates the locations of tumors, it also allows researchers to distinguish between malignant tumors and benign tumors.
To conduct this research project, the scientists analyzed data from six patients identified as having breast tumors with mammograms (X-rays.) In procedures conducted by New York research collaborators Drs. Wei Huang, Alina Tudorica, and Thomas Yankeelov of Stony Brook University and Brookhaven National Laboratory, the patients were injected with a contrast agent, which acts like an MRI dye and provides clearer images. The patients received MRI scans as the dye passed through the tumors. The time courses of the MRI signals were analyzed with the shutter speed model. The results showed hot spots only in images of malignant tumors but not in the benign tumors (three of the cases). This complete distinction was not the case using the standard MRI technique, and there was no distinction using mammography. Pathology results on these tumors confirmed the accuracy of the new MRI testing.
"While continued research is required, we believe shutter speed analyzed MRI could become a powerful tool for the diagnosis and treatment of breast cancer and almost any other form of cancer, as well as many other pathologies," explained Springer. "The shutter speed is a very general concept and applies to a great many different MRI techniques."
"We are fortunate to have recruited Dr. Springer and his team to lead the imaging research activities at OHSU and the OHSU Cancer Institute." said Grover C. Bagby Jr., M.D., Director of the OHSU Cancer Institute. "His 'shutter-speed' model has the potential of changing our approach to cancer screening in general and may also play a role in determining the early effects of treatment. The findings also provide a unique opportunity for cancer researchers to unravel the basic molecular causes of the different image signatures."