A combination of real-world evidence and periodic reporting to the FDA should be the cornerstone of how the agency regulates machine learning and artificial intelligence used in software as a medical device (AI/ML-SaMD), Philips said in a comment on the agency’s draft discussion document.

The FDA is wrestling with how it will keep track of ongoing changes to such software products and is considering focused reviews to ensure that any changes are safe and effective, without requiring an entirely new submission for each iteration of the software.

GE Healthcare said the FDA should consider an expedited pathway for changes and a  focused review should be limited to the proposed changes and to any new risks.

GE welcomed the FDA’s proposed total product lifecycle approach saying it would allow for more efficient regulatory oversight because it “shares concepts” with the agency’s software pre-certification program.