Performance of large language model clinical data extraction by data domain: A rapid systematic review
The Petauri Evidence team will be presenting this research at ISPOR Europe in Glasgow, in Poster Session 4 on Tuesday 11th November.
In this short video, Emily Hardy (Consultant – Systematic Review, Petauri Evidence) introduces the research, along with their poster ‘Traditional versus generative AI: A rapid systematic review assessing accuracy and efficiency of AI in title/abstract screening’
Poster introduction:
Large language models (LLMs) have the potential to increase efficiencies relative to manually conducted systematic reviews; however, caution remains to ensure gold standards are not compromised.
Research objectives:
- Collate studies reporting LLM data extraction of clinical publications
- Explore performance of LLM extraction according to data domain
- Identify any factors influencing LLM extraction performance
Please complete the form below to download the full poster:
Please see the PRISMA and list of included studies for this poster here: https://petauri.com/slr-llm-doc/
Back to ISPOR Europe Virtual Booth
To help save the planet and to save you from carrying a tonne of literature around the exhibition hall, all our resources are available virtually.