We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CL

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computation and Language

Title: Natural Language Processing Methods to Identify Oncology Patients at High Risk for Acute Care with Clinical Notes

Abstract: Clinical notes are an essential component of a health record. This paper evaluates how natural language processing (NLP) can be used to identify the risk of acute care use (ACU) in oncology patients, once chemotherapy starts. Risk prediction using structured health data (SHD) is now standard, but predictions using free-text formats are complex. This paper explores the use of free-text notes for the prediction of ACU instead of SHD. Deep Learning models were compared to manually engineered language features. Results show that SHD models minimally outperform NLP models; an l1-penalised logistic regression with SHD achieved a C-statistic of 0.748 (95%-CI: 0.735, 0.762), while the same model with language features achieved 0.730 (95%-CI: 0.717, 0.745) and a transformer-based model achieved 0.702 (95%-CI: 0.688, 0.717). This paper shows how language models can be used in clinical applications and underlines how risk bias is different for diverse patient groups, even using only free-text data.
Comments: 11 pages, 6 figures, 2 tables
Subjects: Computation and Language (cs.CL); Machine Learning (cs.LG)
Cite as: arXiv:2209.13860 [cs.CL]
  (or arXiv:2209.13860v1 [cs.CL] for this version)

Submission history

From: Claudio Fanconi [view email]
[v1] Wed, 28 Sep 2022 06:31:19 GMT (620kb,D)

Link back to: arXiv, form interface, contact.