"Would you trust an algorithm to help you with a medical diagnosis?" This question is posed by Christina Farr of Fast Company in her discussion of a collaboration between University of California, San Francisco (UCSF) and General Electric with the goal of finding out what Big Data approaches to diagnosis can – and can't – accomplish. The two organizations will be partnering for the next three years to "develop a set of algorithms to help radiologists distinguish between a normal result and one that requires further attention." Knowing the medical community will be skeptical about such machine-learning approaches, not to mention the lack of appropriate regulation for a diagnosis by a non-human, Michael Blum of UCSF notes, "There is a lot of concern from the public and from clinicians that we’ll be developing things to replace doctors. These developments will be focused on supporting clinicians and in developing safer workflows." Read the entire article here.
Friday, November 18, 2016