Jie Liu was awarded the Marco Ramoni Distinguished Paper Award from the American Medical Informatics Association (AMIA) for a paper entitled “New Genetic Variants Improve Personalized Breast Cancer Diagnosis.” Liu, a graduate student co-mentored by Beth Burnside, M.D., M.Ph., Associate Professor (Tenure), Vice Chair of Research, and David Page, Ph.D., has spent his doctoral studies leveraging clinical data sets containing imaging and breast cancer data to develop and test novel machine learning algorithms and statistical methods in the hopes of improving care. Jie presented his award-winning work at the AMIA Summit on Translational Bioinformatics in San Francisco, CA on April 8th, 2014.
This award recognizes a project that was originally supported in 2009-2010 by both the state of Wisconsin, through the Wisconsin Genomics Initiative, and the NIH, though stimulus funds. This project merges mammography features and germline genetic variants in order to better predict breast cancer risk as well as discover novel biologic mechanisms that may be reflected in intermediate phenotypes seen on imaging. In this specific paper, the project team demonstrated that mammography features outperformed clinical and genetic variables in predicting benign and malignant disease in women undergoing breast biopsy. In addition, when imaging, genetic and clinical variables were combined, each data type augmented discriminative performance.
Drs. Burnside and Page maintain an ongoing collaboration with the Marshfield Clinic, which leverages the Personalized Medicine Research Project (PMRP) established in 2002. This population-based initiative created a data repository with stored DNA, plasma and serum; data from surveys; and clinical variables from a comprehensive electronic medical record. The PMRP database allows multi-disciplinary teams of scientists to conduct research in the areas of genetic basis of disease, pharmacogenetics, and population genetics. As genetic disease prediction gains momentum, showing that imaging abnormality features alone outperform and together augment demographic and genetic risk factors in the prediction of breast cancer is crucially important to demonstrate the value of imaging and advance precision medicine.
Paper Details: Liu J, Page D, Peissig P, McCarty C, Onitilo AA, Trentham-Dietz A, Burnside ES. New Genetic Variants Improve Personalized Breast Cancer Diagnosis. AMIA 2014 Joint Summits on Translational Science, SanFrancisco, CA April 8, 2014