乳腺扫描人工智能算法与临床风险模型对5年乳腺癌风险预测的比较:一项观察性研究

Vignesh A Arasu¹, Laurel A Habel¹, Ninah S Achacoso¹+17

• ¹From the Division of Research, Kaiser Permanente Northern California, 2000 Broadway, Oakland, CA 94612 (V.A.A., L.A.H., N.S.A., L.H.K., V.X.L., C.M.L., C.L.); Department of Radiology, Kaiser Permanente Northern California, Vallejo Medical Center, Vallejo, Calif (V.A.A.); Kaiser Permanente Washington Health Research Institute, Seattle, Wash (D.S.M.B.); Department of Radiology, Southern California Permanente Medical Group, Orange County, Irvine, Calif (J.B.C.); Department of Surgery (L.J.E.), Department of Radiology and Biomedical Imaging (N.M.H.), and Department of Epidemiology and Biostatistics (M.M.G., J.K.), University of California-San Francisco, San Francisco, Calif; Department of Medical Imaging Technology and Informatics, Southern California Permanente Medical Group, Pasadena, Calif (D.A.L.); Department of Biostatistics, University of California-Davis, Davis, Calif (D.L.M.); The Technology Group, The Permanente Medical Group, Oakland, Calif (D.A.N.); KP Information Technology, Kaiser Foundation Health Plan Inc and Kaiser Foundation Hospitals, Oakland, Calif (A.P.); Department of Artificial Intelligence and Human Health and Nash Family Department of Neuroscience (L.S.) and Department of Population Health Science and Policy, Department of Genetics and Genomic Sciences (W.S.), Icahn School of Medicine at Mount Sinai, New York, NY; and Department of Radiology, Hawaii Permanente Medical Group, Moanalua Medical Center, Honolulu, Hawaii (H.C.Y.).