AI has captured much attention in medical practice, particularly as some famously predicted that it would replace radiologists in just a few years. While that prediction is unlikely to come true, AI will significantly impact how radiology and medicine are practiced. This lecture will briefly describe some of the technical aspects of AI for medical imaging, proceeding on to discuss some of the applications of AI to medical imaging. Attention will then be paid to the challenges in applying AI, including data issues, regulatory issues, and implementation issues.
Bradley J Erickson, MD PhD, received his MD and PhD degrees from Mayo Clinic. He went on to be trained in radiology, and then completed a Neuroradiology fellowship at Mayo, and has been on staff at Mayo since 1994. He continues to practice Neuroradiology, but has also served in other roles, including chair of the Radiology Informatics Division, Associate Chair for Research, and the Scientific Director for AI. He now serves as the Director for AI at Mayo Clinic, focusing on helping clinical areas develop and implement AI tools into their practices. He has been awarded multiple external grants, including NIH grants on MS, brain tumors, polycystic kidney disease, and medical image exchange. He is a former president of the Society of Imaging Informatics in Medicine, was Chair of the American Board of Imaging Informatics, was on the Board of the IHE USA and chairs the External Advisory Board of the Imaging Data Commons. He recently won the NVIDIA Global Impact Award for his work on deep learning applications in medical images.