Description
Artificial intelligence's vast potential in biomedicine demands rigorous research with comprehensive method and data details. Despite journal peer review and checklists, variations in validation, metrics, data/code accessibility persist in publications, hampering comparability and reproducibility. The AIMe registry, a community-driven platform, enhances biomedical AI model accessibility, reproducibility, and usability. Numerous examples illustrate the critical need for comprehensive reporting in biomedical research to avoid misleading claims. An extensive yet adaptable questionnaire (at https://aime-registry.org) can significantly enhance research documentation and reproducibility in this field.
Keywords
accessibility, reproducibility, usability of biomedical AI models
| Find me @ my poster | 114 |
|---|
