Assigning ICD-O-3 codes to pathology reports using neural multi-task training with hierarchical regularization

Anthony Rios, Eric B. Durbin, Isaac Hands, Ramakanth Kavuluru

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Scopus citations

Abstract

Tracking population-level cancer information is essential for researchers, clinicians, policymakers, and the public. Unfortunately, much of the information is stored as unstructured data in pathology reports. Thus, too process the information, we require either automated extraction techniques or manual curation. Moreover, many of the cancer-related concepts appear infrequently in real-world training datasets. Automated extraction is difficult because of the limited data. This study introduces a novel technique that incorporates structured expert knowledge to improve histology and topography code classification models. Using pathology reports collected from the Kentucky Cancer Registry, we introduce a novel multi-task training approach with hierarchical regularization that incorporates structured information about the International Classification of Diseases for Oncology, 3rd Edition classes to improve predictive performance. Overall, we find that our method improves both micro and macro F1. For macro F1, we achieve up to a 6% absolute improvement for topography codes and up to 4% absolute improvement for histology codes.

Original languageEnglish
Title of host publicationProceedings of the 12th ACM Conference on Bioinformatics, Computational Biology, and Health Informatics, BCB 2021
ISBN (Electronic)9781450384506
DOIs
StatePublished - Jan 18 2021
Event12th ACM Conference on Bioinformatics, Computational Biology, and Health Informatics, BCB 2021 - Virtual, Online, United States
Duration: Aug 1 2021Aug 4 2021

Publication series

NameProceedings of the 12th ACM Conference on Bioinformatics, Computational Biology, and Health Informatics, BCB 2021

Conference

Conference12th ACM Conference on Bioinformatics, Computational Biology, and Health Informatics, BCB 2021
Country/TerritoryUnited States
CityVirtual, Online
Period8/1/218/4/21

Bibliographical note

Publisher Copyright:
© 2021 ACM.

Keywords

  • biomedical informatics
  • natural language processing
  • neural networks

ASJC Scopus subject areas

  • Computer Science Applications
  • Software
  • Biomedical Engineering
  • Health Informatics

Fingerprint

Dive into the research topics of 'Assigning ICD-O-3 codes to pathology reports using neural multi-task training with hierarchical regularization'. Together they form a unique fingerprint.

Cite this