Challenges in understanding clinical notes: Why NLP engines fall short and where background knowledge can help

Sujan Perera; Amit Sheth; Krishnaprasad Thirunarayan; Suhas Nair; Neil Shah

doi:10.1145/2512410.2512427

Challenges in understanding clinical notes: Why NLP engines fall short and where background knowledge can help

Sujan Perera, Amit Sheth, Krishnaprasad Thirunarayan, Suhas Nair, Neil Shah

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Understanding of Electronic Medical Records(EMRs) plays a crucial role in improving healthcare outcomes. However, the unstructured nature of EMRs poses several technical challenges for structured information extraction from clinical notes leading to automatic analysis. Natural Language Processing(NLP) techniques developed to process EMRs are effective for variety of tasks, they often fail to preserve the semantics of original information expressed in EMRs, particularly in complex scenarios. This paper illustrates the complexity of the problems involved and deals with conflicts created due to the shortcomings of NLP techniques and demonstrates where domain specific knowledge bases can come to rescue in resolving conflicts that can significantly improve the semantic annotation and structured information extraction. We discuss various insights gained from our study on real world dataset.

Original language	American English
Title of host publication	DARE '13
Subtitle of host publication	Proceedings of the 2013 international workshop on Data management & analytics for healthcare
Pages	21-26
Number of pages	6
DOIs	https://doi.org/10.1145/2512410.2512427
State	Published - Nov 1 2013
Event	2013 International Workshop on Data Management and Analytics for Healthcare - San Francisco, CA, United States Duration: Nov 1 2013 → Nov 1 2013

Conference

Conference	2013 International Workshop on Data Management and Analytics for Healthcare
Abbreviated title	DARE 2013
Country/Territory	United States
City	San Francisco, CA
Period	11/1/13 → 11/1/13
Other	Co-located with the 22nd ACM International Conference on Information and Knowledge Management, CIKM 2013

ASJC Scopus Subject Areas

General Decision Sciences
General Business,Management and Accounting

Keywords

Knowledge base
Natural language processing
Negation detection

Disciplines

Bioinformatics
Communication
Communication Technology and New Media
Computer Sciences
Databases and Information Systems
Life Sciences
OS and Networks
Physical Sciences and Mathematics
Science and Technology Studies

Access to Document

10.1145/2512410.2512427License: Unspecified

Cite this

Challenges in understanding clinical notes: Why NLP engines fall short and where background knowledge can help. / Perera, Sujan; Sheth, Amit; Thirunarayan, Krishnaprasad et al.
DARE '13: Proceedings of the 2013 international workshop on Data management & analytics for healthcare. 2013. p. 21-26.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Perera, S, Sheth, A, Thirunarayan, K, Nair, S & Shah, N 2013, Challenges in understanding clinical notes: Why NLP engines fall short and where background knowledge can help. in DARE '13: Proceedings of the 2013 international workshop on Data management & analytics for healthcare. pp. 21-26, 2013 International Workshop on Data Management and Analytics for Healthcare, San Francisco, CA, United States, 11/1/13. https://doi.org/10.1145/2512410.2512427

@inproceedings{96fe3ed78e46427eb7c328f5b976cb6e,

title = "Challenges in understanding clinical notes: Why NLP engines fall short and where background knowledge can help",

abstract = " Understanding of Electronic Medical Records(EMRs) plays a crucial role in improving healthcare outcomes. However, the unstructured nature of EMRs poses several technical challenges for structured information extraction from clinical notes leading to automatic analysis. Natural Language Processing(NLP) techniques developed to process EMRs are effective for variety of tasks, they often fail to preserve the semantics of original information expressed in EMRs, particularly in complex scenarios. This paper illustrates the complexity of the problems involved and deals with conflicts created due to the shortcomings of NLP techniques and demonstrates where domain specific knowledge bases can come to rescue in resolving conflicts that can significantly improve the semantic annotation and structured information extraction. We discuss various insights gained from our study on real world dataset.",

keywords = "Knowledge base, Natural language processing, Negation detection",

author = "Sujan Perera and Amit Sheth and Krishnaprasad Thirunarayan and Suhas Nair and Neil Shah",

year = "2013",

month = nov,

day = "1",

doi = "10.1145/2512410.2512427",

language = "American English",

isbn = "9781450324250",

pages = "21--26",

booktitle = "DARE '13",

note = "2013 International Workshop on Data Management and Analytics for Healthcare, DARE 2013 ; Conference date: 01-11-2013 Through 01-11-2013",

}

TY - GEN

T1 - Challenges in understanding clinical notes

T2 - 2013 International Workshop on Data Management and Analytics for Healthcare

AU - Perera, Sujan

AU - Sheth, Amit

AU - Thirunarayan, Krishnaprasad

AU - Nair, Suhas

AU - Shah, Neil

PY - 2013/11/1

Y1 - 2013/11/1

N2 - Understanding of Electronic Medical Records(EMRs) plays a crucial role in improving healthcare outcomes. However, the unstructured nature of EMRs poses several technical challenges for structured information extraction from clinical notes leading to automatic analysis. Natural Language Processing(NLP) techniques developed to process EMRs are effective for variety of tasks, they often fail to preserve the semantics of original information expressed in EMRs, particularly in complex scenarios. This paper illustrates the complexity of the problems involved and deals with conflicts created due to the shortcomings of NLP techniques and demonstrates where domain specific knowledge bases can come to rescue in resolving conflicts that can significantly improve the semantic annotation and structured information extraction. We discuss various insights gained from our study on real world dataset.

AB - Understanding of Electronic Medical Records(EMRs) plays a crucial role in improving healthcare outcomes. However, the unstructured nature of EMRs poses several technical challenges for structured information extraction from clinical notes leading to automatic analysis. Natural Language Processing(NLP) techniques developed to process EMRs are effective for variety of tasks, they often fail to preserve the semantics of original information expressed in EMRs, particularly in complex scenarios. This paper illustrates the complexity of the problems involved and deals with conflicts created due to the shortcomings of NLP techniques and demonstrates where domain specific knowledge bases can come to rescue in resolving conflicts that can significantly improve the semantic annotation and structured information extraction. We discuss various insights gained from our study on real world dataset.

KW - Knowledge base

KW - Natural language processing

KW - Negation detection

UR - http://www.scopus.com/inward/record.url?scp=84889583032&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84889583032&partnerID=8YFLogxK

UR - https://corescholar.libraries.wright.edu/knoesis/561

U2 - 10.1145/2512410.2512427

DO - 10.1145/2512410.2512427

M3 - Conference contribution

SN - 9781450324250

SP - 21

EP - 26

BT - DARE '13

Y2 - 1 November 2013 through 1 November 2013

ER -

Challenges in understanding clinical notes: Why NLP engines fall short and where background knowledge can help

Abstract

Conference

ASJC Scopus Subject Areas

Keywords

Disciplines

Access to Document

Other files and links

Cite this