Development of message passing-based graph convolutional networks for classifying cancer pathology reports
Our implementation revealed that our FastMPN model, which is based on the PyTorch platform, can train a large corpus (667,290 training samples) with 202,373 unique words in less than 3 minutes per epoch using one NVIDIA V100 hardware accelerator. Our experiments demonstrated that using this implementation, the clinical task performance scores of information extraction related to tumors from cancer pathology reports were highly competitive.
Authors
- Antoinette M. Stroup (ORCID: https://orcid.org/0000-0003-3341-4018)
- Georgia D. Tourassi (ORCID: https://orcid.org/0000-0002-9418-9638)
- Andrew E. Blanchard (ORCID: https://orcid.org/0000-0001-7330-9656)
- Hilda Klasky (ORCID: https://orcid.org/0000-0001-7235-2521)
- J. Blair Christian (ORCID: https://orcid.org/0000-0002-4658-1635)
- Eric B. Durbin (ORCID: https://orcid.org/0000-0002-5600-7645)
- Hong‐Jun Yoon (ORCID: https://orcid.org/0000-0002-5450-5878)
- Linda Coyle
- Jennifer A. Doherty (ORCID: https://orcid.org/0000-0002-1454-8187)
- Lynne Penberthy (ORCID: https://orcid.org/0000-0002-9961-8139)
- Xiao-Cheng Wu
Institutions
- Rutgers, The State University of New Jersey (US)
- Oak Ridge National Laboratory (US)
- University of Kentucky (US)
- Huntsman Cancer Institute (US)
- Information Management Services (US)
- National Cancer Institute (US)
- Louisiana State University Health Sciences Center New Orleans (US)
Publication Details
- Journal
- BMC Medical Informatics and Decision Making
- Published
- 2024-09-17
- DOI
- https://doi.org/10.1186/s12911-024-02662-5
- Primary Topic
- Topic Modeling
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- U.S. Department of Energy
- Office of Science
- National Nuclear Security Administration