I am an Associate Professor of Radiology, Electrical and Computer Engineering, and Biostatistics and Bioinformatics at Duke University. I am also the Scientific Director of the Duke Center for Artificial Intelligence in Radiology.
My main research focus is development of machine learning algorithms and their application in medical imaging. We work on a wide range of applications, including breast cancer, brain cancer, thyroid cancer, musculoskeletal imaging and other. Our recent technical work in algorithm development and evaluation includes 3D pyramid pooling in convolutional neural networks, style transfer for image harmonization, and class imbalance in deep learning.
Some of the contributions of my lab have been in the area of imaging-based cancer biomarkers and radiogenomics. Specifically, we showed that algorithmically-extracted imaging features can be used to predict patient outcomes and tumor genomics in breast and brain cancers.
The general goal of my lab is to combine rigorous algorithm development with the highest significance of the solved problems. Toward this goal, the lab members represent both, scientific and medical tracks and have included graduate and undergraduate students, postdoctoral associates, medical residents and fellows, medical students, and visiting scholars. We also work with physicians representing different medical specialties including radiology, medical oncology, surgical oncology, orthopedic surgery, and pediatrics.
I share my work in invited talks in academic and governmental agency settings. I also contribute my more general/philosophical thoughts on the future and ethics of AI in medicine in opinion papers for scientific journals.
My work has been sponsored by the National Institutes of Health, The US Deparment of Defense, Radiological Society of North America, industry, and Duke internal competitive grants.