@VincentRK Metrics for ethics: https://t.co/tWFUVopC7D And https://t.co/Q43nuEmfMK On limitations of medical ethics wrt to AI& patient outcomes: https://t.co/6Agli0vkBx
New paper in Nat Dig Med: https://t.co/GaTpGNQU1a Bias in AI, especially medical autonomous AI, is of great concern, and requires mitigation. Most will be familiar with AI bias from imbalances in the training data, or even incorrect training data. But …
Hi Zak, thanks for addressing. For your readers the following may be of interest: a) with US FDA, we worked out metrics for the quality of reference standards for clinical tasks, see https://t.co/iydXUQvLoC b) IMO medical ethics for AI should focus on …
@DavidEl97019927 @pranavrajpurkar @ZachHarned Fair enough, though there is a whole body of work behind this; ethical foundation https://t.co/211ZNQTnqY Safety and clinical outcomes https://t.co/HTowH1CcD8 Equity & bias https://t.co/vxxVEY75It E…