Specializing natural language processing for the automated clinical coding of patient records
Files
(Published version)
Date
2023
Authors
Nath, Namrata
Editors
Advisors
Journal Title
Journal ISSN
Volume Title
Type:
thesis
Citation
Statement of Responsibility
Conference Name
Abstract
The research presented in this thesis explores ways to enhance Natural Language Processing techniques from the ground up in order to make them better suited to process clinical text. Two fundamental building blocks are developed for the better representation and mining of clinical text. These blocks are subsequently used to address the problem of clinical coding - the process of assigning standardized codes based on discharge notes to assist with billing, analytics and subsequent decision making. The newly developed method of coding is unsupervised, i.e., it does not need to be trained on labelled coding data. Unlike contemporary deep learning models, this method also offers a high degree of interpretability, being able to exactly pinpoint the text snippet that lead to the assignment of a clinical code.
School/Discipline
University of South Australia. UniSA STEM.
UniSA STEM
UniSA STEM
Dissertation Note
Thesis (PhD(Computer and Information Science))--University of South Australia, 2023.
Provenance
Copyright 2023 Namrata Nath.
Description
1 ethesis (xviii, 174 pages) :
colour illustrations.
Includes bibliographical references (pages 127-142)
colour illustrations.
Includes bibliographical references (pages 127-142)
Access Status
506 0#$fstar $2Unrestricted online access