Nature editorial calls for scientific evidence validating medical AI
This editorial emphasizes the importance of rigorous, continuous evidence to validate AI's clinical use, warning against premature adoption without clear standards and comprehensive evaluation guidelines.
A recent editorial from Nature Medicine is calling for sound evidence validating the clinical use of artificial intelligence (AI).
They point out rising advocacy for its use, “even though there is no clear agreement on what level of evidence should be required before such claims are considered credible.” They warn against adopting medical AI too early.
Additional points:
Continuous evidence is needed.
Evaluation guidelines should be written, taking clinical impact into account rather than just “statistical metrics.”
AI validation should be held to the same standard as drug development.
Current regulatory frameworks are “inadequate.”
“Claims of workflow benefit should require implementation studies showing that tools can be integrated without the introduction of delay, burden or unintended harms.”
