Written by : Nikita Saha
December 26, 2024
The recommendations were developed by STANDING Together (STANdards for data Diversity, INclusivity, and Generalisability), an initiative comprising over 350 experts from 58 countries.
A group of international experts has proposed a set of guidelines to minimize bias in artificial intelligence (AI) health technologies.
Published in The Lancet Digital Health and New England Journal of Medicine AI, the recommendations aim to improve the fairness and effectiveness of AI-driven medical innovations across diverse populations.
AI health technologies have been found to produce uneven results, benefiting some groups while potentially disadvantaging or harming others. The recommendations were developed by STANDING Together (STANdards for data Diversity, INclusivity, and Generalisability), an initiative comprising over 350 experts from 58 countries.
“Data is like a mirror, providing a reflection of reality. And when distorted, data can magnify societal biases. But trying to fix the data to fix the problem is like wiping the mirror to remove a stain on your shirt. To create lasting change in health equity, we must focus on fixing the source, not just the reflection,” said Xiaoxuan Liu, an associate professor at the University of Birmingham, UK, and lead author of the guidelines.
The recommendations call for detailed dataset summaries in plain language, identification of biases and errors, and comparative performance evaluations across different demographic groups.
Further, experts also emphasize the importance of mitigation plans for addressing uncertainties in AI performance and transparent reporting of data limitations.
"We hope to raise awareness that no dataset is free of limitations, so transparent communication of data limitations should be perceived as valuable, and absence of this information as a limitation," the experts stated.
According to the authors, these guidelines would promote the development of safe and effective AI technologies.
"We hope that adoption of the STANDING Together recommendations by stakeholders across the AI health technology lifecycle will enable everyone in society to benefit from technologies that are safe and effective," they added.