Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Agarwal, Shubham, Dinu, Vlad, Searle, Thomas, Ratas, Mart, Shek, Anthony, Stein, Dan F., Teo, James, Dobson, Richard
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2501.16077
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866913666977234944
author Agarwal, Shubham
Dinu, Vlad
Searle, Thomas
Ratas, Mart
Shek, Anthony
Stein, Dan F.
Teo, James
Dobson, Richard
author_facet Agarwal, Shubham
Dinu, Vlad
Searle, Thomas
Ratas, Mart
Shek, Anthony
Stein, Dan F.
Teo, James
Dobson, Richard
contents This study introduces RelCAT (Relation Concept Annotation Toolkit), an interactive tool, library, and workflow designed to classify relations between entities extracted from clinical narratives. Building upon the CogStack MedCAT framework, RelCAT addresses the challenge of capturing complete clinical relations dispersed within text. The toolkit implements state-of-the-art machine learning models such as BERT and Llama along with proven evaluation and training methods. We demonstrate a dataset annotation tool (built within MedCATTrainer), model training, and evaluate our methodology on both openly available gold-standard and real-world UK National Health Service (NHS) hospital clinical datasets. We perform extensive experimentation and a comparative analysis of the various publicly available models with varied approaches selected for model fine-tuning. Finally, we achieve macro F1-scores of 0.977 on the gold-standard n2c2, surpassing the previous state-of-the-art performance, and achieve performance of >=0.93 F1 on our NHS gathered datasets.
format Preprint
id arxiv_https___arxiv_org_abs_2501_16077
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle RelCAT: Advancing Extraction of Clinical Inter-Entity Relationships from Unstructured Electronic Health Records
Agarwal, Shubham
Dinu, Vlad
Searle, Thomas
Ratas, Mart
Shek, Anthony
Stein, Dan F.
Teo, James
Dobson, Richard
Computation and Language
This study introduces RelCAT (Relation Concept Annotation Toolkit), an interactive tool, library, and workflow designed to classify relations between entities extracted from clinical narratives. Building upon the CogStack MedCAT framework, RelCAT addresses the challenge of capturing complete clinical relations dispersed within text. The toolkit implements state-of-the-art machine learning models such as BERT and Llama along with proven evaluation and training methods. We demonstrate a dataset annotation tool (built within MedCATTrainer), model training, and evaluate our methodology on both openly available gold-standard and real-world UK National Health Service (NHS) hospital clinical datasets. We perform extensive experimentation and a comparative analysis of the various publicly available models with varied approaches selected for model fine-tuning. Finally, we achieve macro F1-scores of 0.977 on the gold-standard n2c2, surpassing the previous state-of-the-art performance, and achieve performance of >=0.93 F1 on our NHS gathered datasets.
title RelCAT: Advancing Extraction of Clinical Inter-Entity Relationships from Unstructured Electronic Health Records
topic Computation and Language
url https://arxiv.org/abs/2501.16077