A team of researchers led by the University of Exeter in England put together the analysis to help clinical safety professionals and AI developers evaluate current research on medical AI applications and identify areas of concern.
The analysis focuses on machine learning, a type of AI.
“As [machine learning] matures we suggest a set of short-term and medium-term clinical safety issues that need addressing to bring these systems from laboratory to bedside,” the analysis reads.
Four short-term quality and safety issues for hospitals to consider before deploying AI:
1. Distributional shift
2. Insensitivity to impact
3. Black box decision-making
4. Unsafe failure mode
Two medium-term quality and safety issues for hospitals to consider before deploying AI:
1. Automation complacency
2. Reinforcement of outmoded practice and self-fulfilling predictions
To read the analysis in BMJ Quality & Safety, click here.
More articles on artificial intelligence:
AI can re-identify de-identified health data, study finds
Apple promotes AI chief to exec team: 4 things to know
IBM prototypes AI ‘fingernail sensor’ for disease management
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.