AI Guide for Society in Healthcare
Using AI to support healthcare decisions
The guide, for instance, says that the source of the AI data must be clearly known, the data must have been collected or selected for the purpose it is used for, the limitations and assumptions for that purpose are stated, and the solution has been properly tested in the real world. By being transparent and demonstrating the steps taken to check that the AI is reliable, researchers and developers can help give people confidence about providing their data.
The guide was developed in collaboration with the KAIST Korea Policy Center for the Fourth Industrial Revolution (KAIST KPC4IR), and Sense about Science, a non-profit organisation in the UK specialising in science communication. The project team interviewed over 30 experts and practitioners working on AI and Healthcare, and feedback sessions to solicit input from the public.
“The guide was written to motivate and equip the public to participate in deliberations surrounding technological risks. This is especially important in today’s world with deep and rapid technological change that will profoundly affect individuals and healthcare service providers. Through increased societal engagement on the subject, the guide aims to help users understand the benefits and limitations of AI, and ask the pertinent questions which would prompt developers to improve the transparency and quality of AI solutions,” said Prof Koh Chan Ghee, Director of IPUR.
Apart from making an introduction to AI in healthcare (including commonly used terms, AI development landscape and its applications), the guide focuses on the reliability of AI applications in the healthcare sector. For instance, the guide states that the source of the AI data must be clearly known, the data must have been collected or selected for the purpose it is used for, the limitations and assumptions for that purpose are stated, and the solution has been properly tested in the real world. By being transparent and demonstrating the steps taken to check that the AI is reliable, researchers and developers can help give people confidence about providing their data, and to adopt the technology.
The guide was presented during a workshop at the 2021 SIG-KDD (Special Interest Group on Knowledge Discovery and Data Mining) Conference on 15 August 2021. The workshop summary can be accessed here.
Access key resources!