Comparing deep learning and concept extraction based methods for patient phenotyping from clinical narratives
This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. In secondary analysis of electronic health records, a crucial ta...
| Main Authors: | , , , , , , , , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Published: |
Public Library of Science
2018
|
| Online Access: | http://hdl.handle.net/1721.1/114939 https://orcid.org/0000-0002-1119-1346 |
| _version_ | 1811070392131911680 |
|---|---|
| author | Gehrmann, Sebastian Li, Yeran Carlson, Eric T. Wu, Joy T. Welt, Jonathan Foote, John Moseley, Edward T. Grant, David W. Tyler, Patrick D. Dernoncourt, Franck Celi, Leo Anthony G. |
| author2 | Massachusetts Institute of Technology. Institute for Medical Engineering & Science |
| author_facet | Massachusetts Institute of Technology. Institute for Medical Engineering & Science Gehrmann, Sebastian Li, Yeran Carlson, Eric T. Wu, Joy T. Welt, Jonathan Foote, John Moseley, Edward T. Grant, David W. Tyler, Patrick D. Dernoncourt, Franck Celi, Leo Anthony G. |
| author_sort | Gehrmann, Sebastian |
| collection | MIT |
| description | This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. In secondary analysis of electronic health records, a crucial task consists in correctly identifying the patient cohort under investigation. In many cases, the most valuable and relevant information for an accurate classification of medical conditions exist only in clinical narratives. Therefore, it is necessary to use natural language processing (NLP) techniques to extract and evaluate these narratives. The most commonly used approach to this problem relies on extracting a number of clinician-defined medical concepts from text and using machine learning techniques to identify whether a particular patient has a certain condition. However, recent advances in deep learning and NLP enable models to learn a rich representation of (medical) language. Convolutional neural networks (CNN) for text classification can augment the existing techniques by leveraging the representation of language to learn which phrases in a text are relevant for a given medical condition. In this work, we compare concept extraction based methods with CNNs and other commonly used models in NLP in ten phenotyping tasks using 1,610 discharge summaries from the MIMIC-III database. We show that CNNs outperform concept extraction based methods in almost all of the tasks, with an improvement in F1-score of up to 26 and up to 7 percentage points in area under the ROC curve (AUC). We additionally assess the interpretability of both approaches by presenting and evaluating methods that calculate and extract the most salient phrases for a prediction. The results indicate that CNNs are a valid alternative to existing approaches in patient phenotyping and cohort identification, and should be further investigated. Moreover, the deep learning approach presented in this paper can be used to assist clinicians during chart review or support the extraction of billing codes from text by identifying and highlighting relevant phrases for various medical conditions. |
| first_indexed | 2024-09-23T08:35:14Z |
| format | Article |
| id | mit-1721.1/114939 |
| institution | Massachusetts Institute of Technology |
| last_indexed | 2024-09-23T08:35:14Z |
| publishDate | 2018 |
| publisher | Public Library of Science |
| record_format | dspace |
| spelling | mit-1721.1/1149392022-09-30T09:47:37Z Comparing deep learning and concept extraction based methods for patient phenotyping from clinical narratives Gehrmann, Sebastian Li, Yeran Carlson, Eric T. Wu, Joy T. Welt, Jonathan Foote, John Moseley, Edward T. Grant, David W. Tyler, Patrick D. Dernoncourt, Franck Celi, Leo Anthony G. Massachusetts Institute of Technology. Institute for Medical Engineering & Science Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science MIT Critical Data (Laboratory) Dernoncourt, Franck Celi, Leo Anthony G. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. In secondary analysis of electronic health records, a crucial task consists in correctly identifying the patient cohort under investigation. In many cases, the most valuable and relevant information for an accurate classification of medical conditions exist only in clinical narratives. Therefore, it is necessary to use natural language processing (NLP) techniques to extract and evaluate these narratives. The most commonly used approach to this problem relies on extracting a number of clinician-defined medical concepts from text and using machine learning techniques to identify whether a particular patient has a certain condition. However, recent advances in deep learning and NLP enable models to learn a rich representation of (medical) language. Convolutional neural networks (CNN) for text classification can augment the existing techniques by leveraging the representation of language to learn which phrases in a text are relevant for a given medical condition. In this work, we compare concept extraction based methods with CNNs and other commonly used models in NLP in ten phenotyping tasks using 1,610 discharge summaries from the MIMIC-III database. We show that CNNs outperform concept extraction based methods in almost all of the tasks, with an improvement in F1-score of up to 26 and up to 7 percentage points in area under the ROC curve (AUC). We additionally assess the interpretability of both approaches by presenting and evaluating methods that calculate and extract the most salient phrases for a prediction. The results indicate that CNNs are a valid alternative to existing approaches in patient phenotyping and cohort identification, and should be further investigated. Moreover, the deep learning approach presented in this paper can be used to assist clinicians during chart review or support the extraction of billing codes from text by identifying and highlighting relevant phrases for various medical conditions. National Institute of Biomedical Imaging and Bioengineering (U.S.) (Grant R01 EB017205-01A1) 2018-04-24T17:50:16Z 2018-04-24T17:50:16Z 2018-02 2017-06 2018-04-20T17:52:56Z Article http://purl.org/eprint/type/JournalArticle 1932-6203 http://hdl.handle.net/1721.1/114939 Gehrmann, Sebastian et al. “Comparing Deep Learning and Concept Extraction Based Methods for Patient Phenotyping from Clinical Narratives.” Edited by Jen-Hsiang Chuang. PLOS ONE 13, 2 (February 2018): e0192360 © 2018 Gehrmann et al https://orcid.org/0000-0002-1119-1346 http://dx.doi.org/10.1371/journal.pone.0192360 PLOS ONE Creative Commons Attribution 4.0 International License https://creativecommons.org/licenses/by/4.0/ application/pdf Public Library of Science PLoS |
| spellingShingle | Gehrmann, Sebastian Li, Yeran Carlson, Eric T. Wu, Joy T. Welt, Jonathan Foote, John Moseley, Edward T. Grant, David W. Tyler, Patrick D. Dernoncourt, Franck Celi, Leo Anthony G. Comparing deep learning and concept extraction based methods for patient phenotyping from clinical narratives |
| title | Comparing deep learning and concept extraction based methods for patient phenotyping from clinical narratives |
| title_full | Comparing deep learning and concept extraction based methods for patient phenotyping from clinical narratives |
| title_fullStr | Comparing deep learning and concept extraction based methods for patient phenotyping from clinical narratives |
| title_full_unstemmed | Comparing deep learning and concept extraction based methods for patient phenotyping from clinical narratives |
| title_short | Comparing deep learning and concept extraction based methods for patient phenotyping from clinical narratives |
| title_sort | comparing deep learning and concept extraction based methods for patient phenotyping from clinical narratives |
| url | http://hdl.handle.net/1721.1/114939 https://orcid.org/0000-0002-1119-1346 |
| work_keys_str_mv | AT gehrmannsebastian comparingdeeplearningandconceptextractionbasedmethodsforpatientphenotypingfromclinicalnarratives AT liyeran comparingdeeplearningandconceptextractionbasedmethodsforpatientphenotypingfromclinicalnarratives AT carlsonerict comparingdeeplearningandconceptextractionbasedmethodsforpatientphenotypingfromclinicalnarratives AT wujoyt comparingdeeplearningandconceptextractionbasedmethodsforpatientphenotypingfromclinicalnarratives AT weltjonathan comparingdeeplearningandconceptextractionbasedmethodsforpatientphenotypingfromclinicalnarratives AT footejohn comparingdeeplearningandconceptextractionbasedmethodsforpatientphenotypingfromclinicalnarratives AT moseleyedwardt comparingdeeplearningandconceptextractionbasedmethodsforpatientphenotypingfromclinicalnarratives AT grantdavidw comparingdeeplearningandconceptextractionbasedmethodsforpatientphenotypingfromclinicalnarratives AT tylerpatrickd comparingdeeplearningandconceptextractionbasedmethodsforpatientphenotypingfromclinicalnarratives AT dernoncourtfranck comparingdeeplearningandconceptextractionbasedmethodsforpatientphenotypingfromclinicalnarratives AT celileoanthonyg comparingdeeplearningandconceptextractionbasedmethodsforpatientphenotypingfromclinicalnarratives |