A compelling new open-access article titled “Categorising Challenges and Solutions Towards Ethical AI in Breast Cancer Treatment: A Rapid Umbrella Review Complemented by Participatory Methods” has been published in the journal AI and Ethics (Springer Nature).
Authored by Carina Dantas, Miriam Cabrita, Luís Midão, Ana Sofia Carvalho, and Elísio Costa, the study presents a timely and structured examination of the ethical dimensions surrounding the application of Artificial Intelligence in breast cancer care.
The research employs a rapid umbrella review methodology, enriched with insights from participatory stakeholder sessions and interviews. Ethical challenges are categorised into four principal domains:
Individual (e.g. lack of training, algorithmic bias)
Technical (e.g. skewed datasets, poor system design)
Organisational (e.g. absence of auditing processes, lack of diversity in development teams)
Societal (e.g. healthcare inequities, regulatory blind spots)
The article advances an ethics-by-design framework, advocating for early and inclusive multi-stakeholder engagement to proactively identify and address potential biases within AI-driven healthcare systems.
This publication aligns closely with the mission of SHINE 2Europe, which champions inclusive, ethical, and socially responsible innovation across Europe. In particular, the article integrates key findings from the RadioVal social innovation sessions — part of a Horizon Europe project dedicated to validating AI tools for breast cancer diagnosis across diverse health systems and populations.
“A robust ethics-by-design framework, informed by wide-ranging multi-stakeholder engagement, is essential for anticipating and mitigating bias in AI-driven healthcare and ensuring a more equitable use of technology in health,” said Carina Dantas.
The full article is freely available in AI and Ethics: Link