๐๐โ๐ ๐ณ๐ถ๐ป๐ฎ๐น๐น๐ ๐ต๐ฒ๐ฟ๐ฒ!ย After 3 years of dedicated work, weโre thrilled to share our landmark article, published by BMJ, tackling one of healthcare AIโs most critical challenges: ๐ฃ๐ถ๐ช๐ญ๐ฅ๐ช๐ฏ๐จ ๐ต๐ณ๐ถ๐ด๐ต.
๐๐จ๐ง๐จ๐ฅ๐-๐๐ is an international guideline for building trustworthy and deployable AI tools in healthcare, co-developed by ๐ญ๐ญ๐ณ ๐ฒ๐ ๐ฝ๐ฒ๐ฟ๐๐ from ๐ฑ๐ฌ ๐ฐ๐ผ๐๐ป๐๐ฟ๐ถ๐ฒ๐.
The guideline comprises:
โขย 6 core principles for trustworthy AI (Fairness, Universality, Traceability, Usability, Robustness, Explainability), to ensure alignment with key ethical standards.
โขย 30 best practices structured across the four main phases of healthcare AI (design, development, validation, deployment), to provide guidance on practical implementation.
โขย Step-by-step recommendations on how to operationalize each best practice, with examples of possible methods and approaches.
โข Expertise spanning data scientists, clinicians, ethicists, social scientists, and regulators, addressing the technical, clinical, ethical, and legal challenges of healthcare AI.
โก๏ธ Read theย ๐ณ๐๐น๐น ๐ฎ๐ฟ๐๐ถ๐ฐ๐น๐ฒ: https://www.bmj.com/content/388/bmj-2024-081554
๐ก More information? Click ‘Join us’ and subscribe for updates on future initiatives on trustworthy and ethical AI.
