CWRU Links

School of
Medicine

Our Stories

Using Image “Fingerprints” to Diagnose Prostate and Brain Cancer

Radiologist grades tumors with non-invasive imaging technique

Using Image “Fingerprints” to Diagnose Prostate and Brain Cancer

Images in 72-year-old man referred for elevated prostate-specific antigen level of 9.87 ng/mL with minimal urinary symptoms. Patient underwent limited MR imaging and targeted biopsy of lesion in left mid prostate. MR fingerprinting map shows corresponding hypointense lesion in left mid prostate (arrow) and NPZ in right hemiprostate.

Prostate cancer is one of the leading causes of cancer death for American men, with 1 in 7 diagnosed at some point in their life. But screening biopsies are not routinely performed without a pre-identified focus of cancer, and these biopsies, performed transrectally, are invasive and carry significant risk. Patients are reluctant to enter surveillance protocols because of the need for repeated biopsies for monitoring disease. Now, advances in diagnostic imaging could greatly reduce invasive procedures when it comes to spotting cancers.

Vikas Gulani, MD, PhD, is a radiologist and cancer researcher who helped develop a novel kind of magnetic resonance (MR) imaging technology that readily identifies diseased tissue. It is non-invasive and highly sensitive, even discovering cancers before symptoms appear. Gulani, associate professor of radiology, urology, and biomedical engineering at Case Western Reserve University School of Medicine, developed the technology alongside Mark Griswold, PhD, professor in the radiology department and fellow Case Comprehensive Cancer Center member, and Nicole Seiberlich, PhD, assistant professor in the biomedical engineering department.

The technology works by integrating data collected during patient scans into a unique “fingerprint” or signature of data signals for each target tissue. The fingerprints are dependent on physical tissue properties of interest and can be matched against a dictionary of all possible fingerprints, providing detailed, quantitative information about the tissue. Aptly named MR Fingerprinting, it offers a major improvement over traditional MRI techniques, and has recently shown promise as a clinical diagnostic tool.

Gulani put the technology to the test in a recent study published in Radiology. Together with Lee Ponsky, MD, professor of urology and member of the Case Comprehensive Cancer Center, and other researchers from University Hospitals, the team used MR Fingerprinting to create detailed maps of the prostate glands of 140 patients at risk for prostate cancer. The approach helped distinguish between low and high grade prostate tumors in the patients, and quantified relevant tumor characteristics—without any invasive procedures. It is one of the first times MR Fingerprinting has been successfully used in a clinical setting.

“This work is particularly exciting, as it shows that we can not only quantitatively characterize tissue by separating cancer from normal tissue through a combination of MR Fingerprinting with other forms of quantitative MRI, but also may have the ability to better determine cancer aggressiveness in this manner,” Gulani said. “This is a long standing goal of prostate imaging.”

Gulani says other clinical applications for the technology are in the works. MR Fingerprinting has been tested on brain tumors, with the help of University Hospitals Neuroradiologist Chaitra Badve, MD and Andrew Sloan, MD of University Hospitals Neurological Surgery. In the American Journal of Neuroradiology, the team showed MR Fingerprinting can distinguish between grades of glioma tumors, and distinguish metastases and primary brain tumors. The preliminary findings offer hope for patients faced with cancers or diseases where other diagnostic techniques, like biopsies, are highly undesirable. Next up, the researchers plan to test MR Fingerprinting’s ability to differentiate residual brain tumors from radiation treatment effects, a common and particularly difficult problem for doctors trying to interpret brain scans.

“Taken together, the two papers represent true translation of MR Fingerprinting into the clinical setting, which is an important step in showing that the technology will have broad impact and can be robustly used in the challenging environment of clinical imaging,” Gulani said. Gulani is currently collaborating with colleagues to investigate whether the technology can also be used to characterize breast, rectal, and liver cancers.

Gulani anticipates MR Fingerprinting will retain its benefits over traditional MR imaging techniques in all kinds of clinical settings. “Today’s MR images typically only show the degree to which one tissue is more affected by a given property than another,” he explained. “The radiologist subjectively makes inferences about the tissues by looking at the anatomy and relative brightness of tissue. The process is slow and qualitative, leading to difficulty in comparing images day to day, scanner to scanner, before and after treatment, etc. There is also considerable variation in interpretation of the images, depending on the skill level/astuteness of the interpreting physician.”  

During MR Fingerprinting, multiple image properties are quantitatively mapped simultaneously. The advanced technology is more objective than traditional MR imaging. “The major difference is the quantitative component. Each pixel in a quantitative map obtained with MR Fingerprinting is a direct measure of a physical property,” Gulani said. “With MR Fingerprinting, we move from depiction of anatomy followed by human subjective tissue characterization followed by interpretation, towards depiction of anatomy plus quantitative tissue characterization with less subjective interpretation.”

Radiologists can also easily repeat MR Fingerprint scans without sacrificing accuracy, allowing doctors to objectively compare maps over time and track disease progression. Over time, “the technology could help sort out the aggressiveness of cancer non-invasively,” Gulani said.

The novel technology opens new paths for deciphering disease pathology, and measuring treatment responses. Gulani has also worked to extend the technology to dynamic body processes, such as the moving liver and kidneys, and Seiberlich has developed MR Fingerprinting of the beating heart, with exciting results. In the future, Gulani hopes to apply machine-based learning strategies to the computer systems that calculate MR Fingerprinting data to further improve their accuracy. 

CWRUMed360 - The School of Medicine Monthly Newsletter

archives
connect with us
Facebook Twitter YouTube Instagram