Authors: Brenda Y. Miao, Irene Y. Chen, Christopher YK Williams, Jaysón Davidson, Augusto Garcia-Agundez, Harry Sun, Travis Zack, Atul J. Butte, Madhumita Sushil

Published on: March 05, 2024

Impact Score: 7.8

Arxiv code: Arxiv:2403.02558

Summary

• What is new: Incorporating advances in generative models, including LLMs, VLMs, and diffusion models into medical research, and updating the MI-CLAIM checklist for these models.
• Why this is important: Scaling and evaluating the use of advanced generative models in medicine presents new challenges not covered by existing guidelines.
• What the research proposes: Modifying the MI-CLAIM checklist to address training, evaluation, interpretability, and reproducibility of generative models, as well as clarifying cohort selection and aligning with ethical standards.
• Results: An updated MI-CLAIM checklist that accommodates the unique characteristics of generative models in clinical AI research, aiming for more transparent, reproducible research.

Technical Details

Technological frameworks used: Large language models (LLMs), vision language models (VLMs), and diffusion models

Models used: Generative models for natural language and image processing

Data used: Clinical and biomedical data

Potential Impact

Healthcare technology companies, medical research institutions, and developers of clinical AI tools

Want to implement this idea in a business?

We have generated a startup concept here: MediGenAI.