EUVIMEDEUVIMED entdecken
Crossref

Using a Longformer Large Language Model for Segmenting Unstructured Cancer Pathology Reports

Damien Fung iD, Gregory Arbour iD, Krisha Malik, Kaitlin Muzio iD, Raymond Ng

JCO Clinical Cancer Informatics · 2025

DOI: 10.1200/cci-24-00143 ↗

Kurz gefasst

Fachartikel in der Zeitschrift JCO Clinical Cancer Informatics, 2025. Verfasst von Damien Fung und 4 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computer Science, University of British Columbia, Vancouver, Canada. Hauptthemen laut MEDLINE: Natural Language Processing, Neoplasms. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 17 Literaturangaben.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

PURPOSE Many Natural Language Processing (NLP) methods achieve greater performance when the input text is preprocessed to remove extraneous or unnecessary text. A technique known as text segmentation can facilitate this step by isolating key sections from a document. Give that transformer models—such as Bidirectional Encoder Representations from Transformers (BERT)—have demonstrated state-of-the-art performance on many NLP tasks, it is desirable to leverage such models for segmentation. However, transformer models are typically limited to only 512 input tokens and are not well suited for lengthy documents such as cancer pathology reports. The Longformer is a modified transformer model designed to intake longer documents while retaining the positive characteristics of standard transformers. This study presents a Longformer model fine-tuned for cancer pathology report segmentation. METHODS We fine-tuned a Longformer Question-Answer (QA) model on 504 manually annotated pathology reports to isolate sections such as diagnosis, addenda, and clinical history. We compared baseline methods including regular expressions (regex) and BERT QA. However, those methods may fail to correctly identify section boundaries. Model performance was evaluated using sequence recall, precision, and F1 score. RESULTS Final test results were obtained on a hold-out test set of 304 cancer pathology reports. We report sequence F1 scores for the following sections: diagnosis (0.77), addenda (0.48), clinical history (0.89), and overall (0.68). CONCLUSION We present a fine-tuned Longformer model to isolate key sections from cancer pathology reports for downstream analyses. Our model performs segmentation with greater accuracy.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Natural Language ProcessingNeoplasmsHumansElectronic Health RecordsAlgorithmsLarge Language Models

Autor:innen und Institutionen

  • Department of Computer Science, University of British Columbia, Vancouver, Canada.
  • Faculty of Health Sciences, University of Waterloo, Waterloo, Canada.

ORCID-Profile: Damien Fung, Gregory Arbour, Kaitlin Muzio

Zitieren

APA

Fung, D., Arbour, G., Malik, K., Muzio, K., & Ng, R. (2025). Using a Longformer Large Language Model for Segmenting Unstructured Cancer Pathology Reports. JCO Clinical Cancer Informatics. https://doi.org/10.1200/cci-24-00143

Vancouver

Fung D, Arbour G, Malik K, Muzio K, Ng R. Using a Longformer Large Language Model for Segmenting Unstructured Cancer Pathology Reports. JCO Clinical Cancer Informatics. 2025. doi:10.1200/cci-24-00143

BibTeX
@article{Fung2025,
  author = {Fung, Damien and Arbour, Gregory and Malik, Krisha and Muzio, Kaitlin and Ng, Raymond},
  title = {Using a Longformer Large Language Model for Segmenting Unstructured Cancer Pathology Reports},
  journal = {JCO Clinical Cancer Informatics},
  year = {2025},
  doi = {10.1200/cci-24-00143}
}
Kostenlos starten →