The automation of bias in medical Artificial Intelligence (AI): Decoding the past to create a better future
作者：
Highlights：
• Medical errors and biases have disadvantaged vulnerable patient groups for centuries.
• These biases persist today in our medical assessment frameworks, diagnostic tools and educational curricula.
• New Medical AI systems built from current clinical practice pose the risk of further widening population health inequalities.
• For AI to account for the past and build a better future, we must first unpack the present to create a new baseline on which to develop these tools.
• The decisions we make now will determine whether we codify and exacerbate existing inequity or whether we reflect on what we hold to be true and challenge ourselves to be better.
摘要
•Medical errors and biases have disadvantaged vulnerable patient groups for centuries.•These biases persist today in our medical assessment frameworks, diagnostic tools and educational curricula.•New Medical AI systems built from current clinical practice pose the risk of further widening population health inequalities.•For AI to account for the past and build a better future, we must first unpack the present to create a new baseline on which to develop these tools.•The decisions we make now will determine whether we codify and exacerbate existing inequity or whether we reflect on what we hold to be true and challenge ourselves to be better.
论文关键词：Disparities,Inequality,Data science,Bias,Health,Medicine,Digital health,Artificial intelligence,Healthcare
论文评审过程：Received 3 April 2020, Revised 28 August 2020, Accepted 1 October 2020, Available online 6 October 2020, Version of Record 14 October 2020.
论文官网地址：https://doi.org/10.1016/j.artmed.2020.101965
