Trustworthiness of artificial intelligence (AI) applications and validating data analysis in cancer care was recently discussed at the ESMO Women for Oncology Forum.
Click here to read about an interview with Alessandra Pedrocchi, founder of the laboratory for oncology at the National Cancer Institute in Milan.
Trust and explainability are the right conditions to put on AI in oncology, because a model that cannot show its reasoning will not be trusted with treatment decisions no matter how accurate it looks in testing. Validating data analysis matters just as much — biased or unrepresentative training data quietly becomes unequal care. This is the kind of implementation question the oncology brothers discussions handle well: separating what AI promises from what it can actually justify in clinic.
ReplyDelete