Bias, Fairness and Generalizability
Description
Dr. Leo Celi discusses various problems involving bias, fairness and generalizability that continue to affect the adoption of artificial intelligence models in hospitals and clinics. Dr. Celi also makes a number of recommendations for improving relationships between health care organizations and the private sector as AI research moves forward.
Articles that Dr. Celi mentions in the podcast are:
Futoma J, Simons M, Panch T, Doshi-Velez F, Celi L. The myth of generalizability in clinical research and machine learning in health care. Lancet Digital Health 2020; 2:e489-92. At: https://www.thelancet.com/journals/landig/article/PIIS2589-7500(20)30186-2/fulltext.
Stuppe A, Singerman D, Celi L. The reproducibility crisis in the age of digital medicine. NPJ Digital Medicine January 29, 2019. At https://www.nature.com/articles/s41746-019-0079-z.
Vyas D, Eisenstein L, Jones D. Hidden in plain sight—reconsidering the use of race correction in clinical algorithms. The New England Journal of Medicine August 27, 2020; 383(9):874-882. At: https://www.nejm.org/doi/full/10.1056/NEJMms2004740.
United States
