Epic founder and CEO Judy Faulkner cautions that AI can be manipulated, so the healthcare industry should cautiously monitor its evolution.
“Lots of things have been gamed lately. Capitalism has been gamed. Government has been gamed. I think AI we have to watch to make sure it doesn’t get gamed,” she told “Freakonomics Radio” for an April 24 episode. “If you want AI to say a certain something, you put it in repeatedly all over the place. It goes back to if you repeat something often enough, people will think it’s true. Well, AI will think it’s true.”
But the technology has enormous upsides, she said on the podcast, which is why Epic is dedicating so much research and development spending to AI. A big benefit: predicting disease. She said a customer recently told Epic that its AI sepsis detection tool cut its cases to zero, saving over 100 lives.
AI can also match like patients, improving the treatment of rare diseases. “If you are a very odd patient with problems that your doctor has never seen before and doesn’t know how to care for you, and has tried this and this and this and nothing’s really working, we can find the two other patients in the United States who are similar to you, and have the three caregivers be able to talk to each other so they can take better care of you,” Ms. Faulkner’s told the podcast’s host, Stephen Dubner.
In addition, the technology can scan massive amounts of anonymized EHR data to help clinicians with diagnostic decisions.
“We may say the 32,500 patients very similar to your patient, 55% of the time this was the diagnosis given, the diagnosis you gave was used 0.2% of the time,” Ms. Faulkner said. “It’s not telling the physician to change anything, but it is giving the physician an alert that others have made different decisions, so the physician can decide if he or she is right or if he or she is wrong.”
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.