Lessons Learned in Building Expertly Annotated Multi-Institution Datasets and Hosting the RSNA AI Challenges

Felipe C Kitamura¹, Luciano M Prevedello¹, Errol Colak¹+14

• ¹From the Department of Applied Innovation and AI, Dasa, São Paulo, Brazil (F.C.K.); Department of Diagnostic Imaging, Universidade Federal de São Paulo (Unifesp), Av Prof Ascendino Reis, 1245, 131, São Paulo, SP, Brazil 04027-000 (F.C.K.); Department of Radiology, The Ohio State University Wexner Medical Center, Columbus, Ohio (L.M.P.); Department of Medical Imaging, University of Toronto, Toronto, Canada (E.C.); Ann and Robert H. Lurie Children's Hospital of Chicago, Chicago, Ill (S.S.H.); Microsoft HLS, Redmond, Wash (M.P.L.); Department of Biomedical Data Science, Stanford University, Stanford, Calif (M.P.L.); The Jackson Laboratory, Bar Harbor, Maine (R.L.B.); Department of Ophthalmology, University of Colorado Denver School of Medicine, Aurora, Colo (J.K.C.); Department of Radiology, University of Pennsylvania, Philadelphia, Pa (C.E.K.); Department of Radiology, University of Utah, Salt Lake City, Utah (T.R.); Department of Radiology and Biomedical Imaging (M.P.L., J.F.T., J.M.) and Center for Intelligent Imaging (J.M.), University of California San Francisco, San Francisco, Calif; Department of Radiology, Weill Cornell Medical College, New York, NY (G.S.); Department of Medical Imaging, Unity Health Toronto, Toronto, Canada (H.M.L.); Department of Radiology, Brigham and Women's Hospital, Harvard Medical School, MGB Data Science Office, Boston, Mass (K.P.A.); Informatics Department, Radiological Society of North America, Oak Brook, Ill (M.V.); Department of Radiology, Mayo Clinic, Rochester, Minn (B.J.E.); and Department of Radiology, Thomas Jefferson University, Philadelphia, Pa (A.E.F.).