Artificial intelligence is prone to overdiagnosis - Cancerworld

The use of artificial intelligence might increase the speed and the consistency of cancer diagnosis, but could also exacerbate the problem of overdiagnosis, according to a perspective article recently published in the New England Journal of Medicine by Adewole Adamson and Gilbert Welch, who suggest that this risk may be mitigated by overcoming the dichotomous classification between "cancer" and "not cancer". Supervised machine learning consists in the generation of decision-making algorithms starting from sets of images that pathologists have categorized as either "cancer" or "not cancer." "The computer system learns by judging its diagnosis against the external standard of pathological interpretation" Adewole Adamson, assistant professor of Internal Medicine at Dell Medical School at the University of Texas, explains. "Reliance on this external standard is problematic, however, since machine learning doesn't solve the central problem associated with cancer diagnosis: the lack of a histopathological gold standard." There is no single right answer to the question: "What constitutes cancer?"