New UChicago Technology Aims to Advance Breast Cancer Diagnoses
Enhancements in analyzing lesions to predict response to therapy could make a difference in how patients are managed.
—Maryellen Giger, PhD
A prototype breast imaging computer workstation developed at The University of Chicago could soon revolutionize the way breast cancer is diagnosed. Radiology Professor Maryellen Giger, PhD, and members of her research team, created an intelligent breast workstation for computer-aided diagnosis (CADx) and quantitative image analysis, that reviews data from multimodality images and helps evaluate and characterize suspicious lesions.
To use the system, images are first loaded from a patient’s mammography, ultrasound, and magnetic resonance (MR) exams. Next, radiologists click on the area of the suspicious lesion in each modality. Within seconds, the computer returns characteristic quantitative data about the lesion in the form of image-based biomarkers, such as size, shape, texture, margin, contrast, and kinetic uptake. The merged output from the unknown case is indicated on a histogram along with merged data from cases of known benign and malignant tumors. Using computer intelligence, the CADx technology gathers quantitative data about the lesion and estimates a probability of malignancy based on a reference library of known cases of benign and malignant tumors.
At the Radiological Society of North America’s (RSNA) annual meeting last December, the UChicago researchers demonstrated the technology in RSNA’s quantitative imaging reading room showcase. Dr. Giger noted that the CADx workstation was met with high interest, especially from people involved in quantitative imaging.
“With this system, the interpretative process, which many radiologists perform manually and visually, can be conducted automatically,” Dr. Giger explained. “People appear to be ready to include more computer analyses into their interpretations.”
The next step is to move from a prototype research phase to actual application in a clinical setting. Dr. Giger said a company called Quantitative Insights is currently negotiating a licensing agreement for the product with UChicagoTech, the university’s office of technology and intellectual property. Quantitative Insights was formed by a team of Chicago Booth MBA students who were finalists in the 2010 New Venture Challenge, an annual competition run by the Booth School of Business and its Polsky Center for Entrepreneurship. Dr. Giger collaborated with the team, and the start-up company is now obtaining expert advice on the commercialization of the workstation via the Chicago Innovation Mentors program.
Dr. Giger, who said she is excited to see the technology take one step closer to clinical implementation, said, “I see CADx translating to the clinical setting in the near future as another ‘test result’ to be used concurrently by radiologists in improving the accuracy of diagnostic output and patient management.”
With success already demonstrated in the areas of detection and diagnosis, Dr. Giger said she plans to now focus on improving prognostic capabilities, in other words, harnessing the technology to determine how advanced a cancer is. “Enhancements in analyzing lesions to predict response to therapy could make a difference in how patients are managed,” Dr. Giger said.
Members of Dr. Maryellen Giger’s lab surround the intelligent breast cancer workstation for CADx and quantitative image analysis. The workstation uses information from mammograms, ultrasounds, and MRIs to help radiologists more accurately diagnose suspicious breast lesions.
(sitting, from left) Li Lan, MS, Maryellen Giger, PhD, Karen Drukker, PhD (standing, from left) Hsien-Chi Kuo, Jyothi Janardanan, MS, Hui Li, PhD, Jeremy Bancroft Brown, Michael Chinander, PhD, Robert Tomek, MSc, Andrew Jamieson, and Claire Salling (not pictured) Neha Bhooshan, PhD, Karla Horsch, PhD, and Umnouy Ponsukcharoen