From the Journal of the American Medical Association at jamanetwork.com.
Here is a link to the paper.
According to the paper, "In this study, SCORE-AI achieved human expert level performance in fully automated interpretation of routine EEGs."
AI's accuracy was assessed by comparing its assessment with those of a panel of human experts. Doesn't this procedure limit AI's measured accuracy to that of humans? To the extent that human experts are inaccurate, aren't AI training procedures guaranteed to produce flawed results?
No comments:
Post a Comment