Automated information extraction from free-text medical documents for stroke key performance indicators: a pilot study

Stephen Bacchi; Sam Gluck; Simon Koblar; Jim Jannes; Timothy Kleinig

doi:10.1111/imj.15678

Automated information extraction from free-text medical documents for stroke key performance indicators: a pilot study

Stephen Bacchi
, Sam Gluck
, Simon Koblar
, Jim Jannes
, Timothy Kleinig

Research output: Contribution to journal › Article › peer-review

8 Citations (Scopus)

Abstract

Automated information extraction might be able to assist with the collection of stroke key performance indicators (KPI). The feasibility of using natural language processing for classification-based KPI and datetime field extraction was assessed. Using free-text discharge summaries, random forest models achieved high levels of performance in classification tasks (area under the receiver operator curve 0.95–1.00). The datetime field extraction method was successful in 29 of 43 (67.4%) cases. Further studies are indicated.

Original language	English
Pages (from-to)	315-317
Number of pages	3
Journal	Internal Medicine Journal
Volume	52
Issue number	2
DOIs	https://doi.org/10.1111/imj.15678
Publication status	Published or Issued - Feb 2022
Externally published	Yes

Keywords

key performance indicator
machine learning
natural language processing
random forest

ASJC Scopus subject areas

Internal Medicine

Access to Document

10.1111/imj.15678

Cite this

@article{ab3b0c75a4044a30a89a5e462a19d2a8,

title = "Automated information extraction from free-text medical documents for stroke key performance indicators: a pilot study",

abstract = "Automated information extraction might be able to assist with the collection of stroke key performance indicators (KPI). The feasibility of using natural language processing for classification-based KPI and datetime field extraction was assessed. Using free-text discharge summaries, random forest models achieved high levels of performance in classification tasks (area under the receiver operator curve 0.95–1.00). The datetime field extraction method was successful in 29 of 43 (67.4\%) cases. Further studies are indicated.",

keywords = "key performance indicator, machine learning, natural language processing, random forest",

author = "Stephen Bacchi and Sam Gluck and Simon Koblar and Jim Jannes and Timothy Kleinig",

note = "Publisher Copyright: {\textcopyright} 2022 Royal Australasian College of Physicians.",

year = "2022",

month = feb,

doi = "10.1111/imj.15678",

language = "English",

volume = "52",

pages = "315--317",

journal = "Internal Medicine Journal",

issn = "1444-0903",

publisher = "John Wiley and Sons Inc.",

number = "2",

}

TY - JOUR

T1 - Automated information extraction from free-text medical documents for stroke key performance indicators

T2 - a pilot study

AU - Bacchi, Stephen

AU - Gluck, Sam

AU - Koblar, Simon

AU - Jannes, Jim

AU - Kleinig, Timothy

PY - 2022/2

Y1 - 2022/2

N2 - Automated information extraction might be able to assist with the collection of stroke key performance indicators (KPI). The feasibility of using natural language processing for classification-based KPI and datetime field extraction was assessed. Using free-text discharge summaries, random forest models achieved high levels of performance in classification tasks (area under the receiver operator curve 0.95–1.00). The datetime field extraction method was successful in 29 of 43 (67.4%) cases. Further studies are indicated.

AB - Automated information extraction might be able to assist with the collection of stroke key performance indicators (KPI). The feasibility of using natural language processing for classification-based KPI and datetime field extraction was assessed. Using free-text discharge summaries, random forest models achieved high levels of performance in classification tasks (area under the receiver operator curve 0.95–1.00). The datetime field extraction method was successful in 29 of 43 (67.4%) cases. Further studies are indicated.

KW - key performance indicator

KW - machine learning

KW - natural language processing

KW - random forest

UR - https://www.scopus.com/pages/publications/85125002831

U2 - 10.1111/imj.15678

DO - 10.1111/imj.15678

M3 - Article

C2 - 35187820

AN - SCOPUS:85125002831

SN - 1444-0903

VL - 52

SP - 315

EP - 317

JO - Internal Medicine Journal

JF - Internal Medicine Journal

IS - 2

ER -

Automated information extraction from free-text medical documents for stroke key performance indicators: a pilot study

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Cite this