Forum for Applied Imaging Research (FAIR): Large Language Model

Dr. Jae Ho Sohn is a cardiothoracic radiologist and co-chair of the education pillar of the UCSF Center for Intelligent Imaging. His research focuses on radiological natural language processing, lung cancer imaging, and clinical translation of 0.55T low-field lung MRI. As a physician with engineering background, he has been broadly involved in projects that use mathematical and engineering techniques to tackle research questions in radiology. His notable work on natural language processing includes automated protocolling of MRI studies, BERT for detection of radiology report dictation errors, and clinical language search algorithm to match clinical indication with appropriate radiological studies, many in collaboration with students, residents, and engineers.

Dr. Yannan Yu is a second year (PGY-3) radiology resident at UCSF Radiology and Biomedical Imaging with plans for T32 research fellowship involving machine learning in radiology, including Large Language Model (LLM).  Dr. Yu completed an internship at University of Massachusetts Medical Center, Worcester, MA in 2022.  She completed her post-doctoral fellowship at Stanford University, Stanford, CA with extensive research focus on computer vision and other machine learning algorithm development in neuroradiology. Dr. Yu received her MD degree in Diagnostic Medicine at Peking Union Medical College, Beijing, China in 2016.