This study introduces HUMANE, a new checklist for researchers and reviewers working on artificial intelligence (AI) and machine learning (ML) in medicine. The goal is to create a common understanding and improve the quality of research in this rapidly growing field.
Why it’s needed:
- AI in healthcare is growing fast
- There’s a big knowledge gap about AI among healthcare professionals
- We lack tools to properly structure and evaluate AI research in medicine
How they did it:
- They used a method called Delphi to get expert opinions
- They created a survey with 8 sections and 56 questions
- Experts rated each item on a scale of 1 to 5
What they found:
- 33 experts participated, mostly from the USA, UK, and Australia
- Most were healthcare professionals in their early careers
- Overall, experts agreed with most of the checklist items
- Some sections were more widely accepted than others
- After getting feedback, they revised the checklist to 8 sections and 50 questions
In conclusion, the HUMANE checklist could be a useful tool to improve the quality and reliability of AI research in medicine. More research is needed to confirm its effectiveness.
Keywords: Artificial intelligence, machine learning, healthcare, medicine, HUMANE checklist, consensus, Delphi method, research guidelines, AI in medicine, medical informatics, digital health, health technology, clinical decision support, predictive modeling, diagnostic algorithms, medical AI, AI ethics, AI transparency, AI explainability, data science, biomedical informatics, clinical informatics, health informatics, AI validation, AI implementation, AI literacy, medical research, research methodology, clinical research, AI bias, AI fairness, AI safety, AI governance, medical imaging AI, natural language processing, clinical NLP, electronic health records, AI standards, AI checklist, AI reporting guidelines, AI quality assessment, AI clinical trials, AI model development, AI model evaluation, AI peer review, medical AI education, AI in clinical practice, AI for precision medicine, AI for personalized medicine, AI for public health.