CPLLM: Clinical prediction with large language models
Ofir Ben Shoham, Nadav Rappoport · 2024
Read CPLLM: Clinical prediction with large language models by Ofir Ben Shoham, Nadav Rappoport (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-11623460.txt.
Description
Ben Shoham and Rappoport present CPLLM, a method that fine-tunes large language models for clinical forecasts. The model reads a patient's past records, written as prompts, to predict future illness. It also forecasts whether a patient will return to hospital.
The team used quantization to shrink the model before fine-tuning it. Results were compared with Retain and Med-BERT, a top model for disease prediction from records. CPLLM beat all tested models on both PR-AUC and ROC-AUC scores. It did so without pre-training on medical data. The authors say it could slot into care workflows to help plan next steps.
Catalog record
| CREATOR | Ofir Ben Shoham, Nadav Rappoport |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 40 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |