Combining Domain-Specific Models and LLMs for Automated Disease Phenotyping from Survey Data

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Beeri, Gal, Chamot, Benoit, Latchem, Elena, Venkatesh, Shruthi, Whalan, Sarah, Kruger, Van Zyl, Martino, David
Format: Preprint
Publié: 2024
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866929641656156160
author Beeri, Gal
Chamot, Benoit
Latchem, Elena
Venkatesh, Shruthi
Whalan, Sarah
Kruger, Van Zyl
Martino, David
author_facet Beeri, Gal
Chamot, Benoit
Latchem, Elena
Venkatesh, Shruthi
Whalan, Sarah
Kruger, Van Zyl
Martino, David
contents This exploratory pilot study investigated the potential of combining a domain-specific model, BERN2, with large language models (LLMs) to enhance automated disease phenotyping from research survey data. Motivated by the need for efficient and accurate methods to harmonize the growing volume of survey data with standardized disease ontologies, we employed BERN2, a biomedical named entity recognition and normalization model, to extract disease information from the ORIGINS birth cohort survey data. After rigorously evaluating BERN2's performance against a manually curated ground truth dataset, we integrated various LLMs using prompt engineering, Retrieval-Augmented Generation (RAG), and Instructional Fine-Tuning (IFT) to refine the model's outputs. BERN2 demonstrated high performance in extracting and normalizing disease mentions, and the integration of LLMs, particularly with Few Shot Inference and RAG orchestration, further improved accuracy. This approach, especially when incorporating structured examples, logical reasoning prompts, and detailed context, offers a promising avenue for developing tools to enable efficient cohort profiling and data harmonization across large, heterogeneous research datasets.
format Preprint
id arxiv_https___arxiv_org_abs_2410_20695
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Combining Domain-Specific Models and LLMs for Automated Disease Phenotyping from Survey Data
Beeri, Gal
Chamot, Benoit
Latchem, Elena
Venkatesh, Shruthi
Whalan, Sarah
Kruger, Van Zyl
Martino, David
Computation and Language
This exploratory pilot study investigated the potential of combining a domain-specific model, BERN2, with large language models (LLMs) to enhance automated disease phenotyping from research survey data. Motivated by the need for efficient and accurate methods to harmonize the growing volume of survey data with standardized disease ontologies, we employed BERN2, a biomedical named entity recognition and normalization model, to extract disease information from the ORIGINS birth cohort survey data. After rigorously evaluating BERN2's performance against a manually curated ground truth dataset, we integrated various LLMs using prompt engineering, Retrieval-Augmented Generation (RAG), and Instructional Fine-Tuning (IFT) to refine the model's outputs. BERN2 demonstrated high performance in extracting and normalizing disease mentions, and the integration of LLMs, particularly with Few Shot Inference and RAG orchestration, further improved accuracy. This approach, especially when incorporating structured examples, logical reasoning prompts, and detailed context, offers a promising avenue for developing tools to enable efficient cohort profiling and data harmonization across large, heterogeneous research datasets.
title Combining Domain-Specific Models and LLMs for Automated Disease Phenotyping from Survey Data
topic Computation and Language
url https://arxiv.org/abs/2410.20695