MedSlice: Fine-Tuned Large Language Models for Secure Clinical Note Sectioning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Davis, Joshua, Sounack, Thomas, Sciacca, Kate, Brain, Jessie M, Durieux, Brigitte N, Agaronnik, Nicole D, Lindvall, Charlotta
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913664148176896
author Davis, Joshua
Sounack, Thomas
Sciacca, Kate
Brain, Jessie M
Durieux, Brigitte N
Agaronnik, Nicole D
Lindvall, Charlotta
author_facet Davis, Joshua
Sounack, Thomas
Sciacca, Kate
Brain, Jessie M
Durieux, Brigitte N
Agaronnik, Nicole D
Lindvall, Charlotta
contents Extracting sections from clinical notes is crucial for downstream analysis but is challenging due to variability in formatting and labor-intensive nature of manual sectioning. While proprietary large language models (LLMs) have shown promise, privacy concerns limit their accessibility. This study develops a pipeline for automated note sectioning using open-source LLMs, focusing on three sections: History of Present Illness, Interval History, and Assessment and Plan. We fine-tuned three open-source LLMs to extract sections using a curated dataset of 487 progress notes, comparing results relative to proprietary models (GPT-4o, GPT-4o mini). Internal and external validity were assessed via precision, recall and F1 score. Fine-tuned Llama 3.1 8B outperformed GPT-4o (F1=0.92). On the external validity test set, performance remained high (F1= 0.85). Fine-tuned open-source LLMs can surpass proprietary models in clinical note sectioning, offering advantages in cost, performance, and accessibility.
format Preprint
id arxiv_https___arxiv_org_abs_2501_14105
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle MedSlice: Fine-Tuned Large Language Models for Secure Clinical Note Sectioning
Davis, Joshua
Sounack, Thomas
Sciacca, Kate
Brain, Jessie M
Durieux, Brigitte N
Agaronnik, Nicole D
Lindvall, Charlotta
Computation and Language
Artificial Intelligence
Information Retrieval
Machine Learning
Extracting sections from clinical notes is crucial for downstream analysis but is challenging due to variability in formatting and labor-intensive nature of manual sectioning. While proprietary large language models (LLMs) have shown promise, privacy concerns limit their accessibility. This study develops a pipeline for automated note sectioning using open-source LLMs, focusing on three sections: History of Present Illness, Interval History, and Assessment and Plan. We fine-tuned three open-source LLMs to extract sections using a curated dataset of 487 progress notes, comparing results relative to proprietary models (GPT-4o, GPT-4o mini). Internal and external validity were assessed via precision, recall and F1 score. Fine-tuned Llama 3.1 8B outperformed GPT-4o (F1=0.92). On the external validity test set, performance remained high (F1= 0.85). Fine-tuned open-source LLMs can surpass proprietary models in clinical note sectioning, offering advantages in cost, performance, and accessibility.
title MedSlice: Fine-Tuned Large Language Models for Secure Clinical Note Sectioning
topic Computation and Language
Artificial Intelligence
Information Retrieval
Machine Learning
url https://arxiv.org/abs/2501.14105