Fine-tuning a 7-billion-parameter language model on UK Biobank patient descriptions yields CVD risk predictions that match gradient-boosted trees on base risk factors and improve when broader health data is added.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Adaptable Cardiovascular Disease Risk Prediction from Heterogeneous Data using Large Language Models
Fine-tuning a 7-billion-parameter language model on UK Biobank patient descriptions yields CVD risk predictions that match gradient-boosted trees on base risk factors and improve when broader health data is added.