Baltimore-based University of Maryland School of Medicine researchers wrote the article to offer guidance on how to implement open-source large language models to radiology departments.
Here are three things to know from the guidance:
- Researchers said there are advantages and disadvantages to using open-source models, such as Llama 3 and LLaVA-Med, in a radiology setting.
Advantages include increased customization, reduced costs, smaller computational requirements and enhanced data security.
Open-source models are still working to “close the gap” on performance and safety metrics, researchers said of potential disadvantages to not using commercial LLMs.
- The article offers instructions on how to select, load and deploy open-source modes within a radiology setting. It also provides implementation-ready code for different open-source tools.
- Clinical deployment considerations for using open-source LLMs include a lack of established best practices due to the rapid development of AI, as well as large language models’ tendency to exhibit biases and perpetuate misinformation, the researchers said.
Read the full guidance here.
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.