DENVER - It's important to test radiology artificial intelligence (AI) algorithms thoroughly with a rigorous methodological approach prior to adoption, said Dr. Charles Kahn of the University of Pennsylvania in a video interview Thursday at the Society for Imaging Informatics in Medicine (SIIM) annual meeting.
In a conversation with AuntMinnie.com Kahn shared a preview of his Friday presentation on how to ensure safe, effective, and humane AI. He will give the talk -- this year's Dwyer lecture -- during SIIM 2019's closing session.
"Part of what we want to do is make sure people understand the importance of testing these systems and testing them in the environment where they are going to be using them and looking for rigor in that process," Kahn said.
Beyond ensuring the safety and effectiveness of radiology AI algorithms, those who are creating this technology -- whether they are physicians, engineers, data scientists, or computer scientists -- also have an obligation to choose how this technology is going to be used, he noted.
"There's tremendous opportunity and potential for AI systems to come into radiology," he said. "I think we just have to be mindful of how we deploy them and how we choose to use them."
