Patient Event Sequences for Predicting Hospitalization Length of Stay

Emil Riis Hansen; Thomas Dyhre Nielsen; Thomas Mulvad; Mads Nibe Strausholm; Tomer Sagi; Katja Hose

doi:10.1007/978-3-031-34344-5_7

Patient Event Sequences for Predicting Hospitalization Length of Stay

Emil Riis Hansen, Thomas Dyhre Nielsen, Thomas Mulvad, Mads Nibe Strausholm, Tomer Sagi, Katja Hose

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

Abstract

Predicting patients’ hospital length of stay (LOS) is essential for improving resource allocation and supporting decision-making in healthcare organizations. This paper proposes a novel transformer-based model, termed Medic-BERT (M-BERT), for predicting LOS by modeling patient information as sequences of events. We performed empirical experiments on a cohort of 48k emergency care patients from a large Danish hospital. Experimental results show that M-BERT can achieve high accuracy on a variety of LOS problems and outperforms traditional non-sequence-based machine learning approaches.

Original language	English
Title of host publication	Artificial Intelligence in Medicine - 21st International Conference on Artificial Intelligence in Medicine, AIME 2023, Proceedings
Editors	Jose M. Juarez, Mar Marcos, Gregor Stiglic, Allan Tucker
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	51-56
Number of pages	6
ISBN (Print)	9783031343438
DOIs	https://doi.org/10.1007/978-3-031-34344-5_7
State	Published - 2023
Externally published	Yes
Event	21st International Conference on Artificial Intelligence in Medicine, AIME 2023 - Portoroz, Slovenia Duration: 12 Jun 2023 → 15 Jun 2023

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	13897 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	21st International Conference on Artificial Intelligence in Medicine, AIME 2023
Country/Territory	Slovenia
City	Portoroz
Period	12/06/23 → 15/06/23

Bibliographical note

Publisher Copyright:
© 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Keywords

length of stay prediction
sequence models
transformers

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-031-34344-5_7

Cite this

Hansen, E. R., Nielsen, T. D., Mulvad, T., Strausholm, M. N., Sagi, T., & Hose, K. (2023). Patient Event Sequences for Predicting Hospitalization Length of Stay. In J. M. Juarez, M. Marcos, G. Stiglic, & A. Tucker (Eds.), Artificial Intelligence in Medicine - 21st International Conference on Artificial Intelligence in Medicine, AIME 2023, Proceedings (pp. 51-56). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13897 LNAI). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-34344-5_7

Hansen, Emil Riis ; Nielsen, Thomas Dyhre ; Mulvad, Thomas et al. / Patient Event Sequences for Predicting Hospitalization Length of Stay. Artificial Intelligence in Medicine - 21st International Conference on Artificial Intelligence in Medicine, AIME 2023, Proceedings. editor / Jose M. Juarez ; Mar Marcos ; Gregor Stiglic ; Allan Tucker. Springer Science and Business Media Deutschland GmbH, 2023. pp. 51-56 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{820d68e949644afebf79064358f2e066,

title = "Patient Event Sequences for Predicting Hospitalization Length of Stay",

abstract = "Predicting patients{\textquoteright} hospital length of stay (LOS) is essential for improving resource allocation and supporting decision-making in healthcare organizations. This paper proposes a novel transformer-based model, termed Medic-BERT (M-BERT), for predicting LOS by modeling patient information as sequences of events. We performed empirical experiments on a cohort of 48k emergency care patients from a large Danish hospital. Experimental results show that M-BERT can achieve high accuracy on a variety of LOS problems and outperforms traditional non-sequence-based machine learning approaches.",

keywords = "length of stay prediction, sequence models, transformers",

author = "Hansen, {Emil Riis} and Nielsen, {Thomas Dyhre} and Thomas Mulvad and Strausholm, {Mads Nibe} and Tomer Sagi and Katja Hose",

note = "Publisher Copyright: {\textcopyright} 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.; 21st International Conference on Artificial Intelligence in Medicine, AIME 2023 ; Conference date: 12-06-2023 Through 15-06-2023",

year = "2023",

doi = "10.1007/978-3-031-34344-5_7",

language = "English",

isbn = "9783031343438",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "51--56",

editor = "Juarez, {Jose M.} and Mar Marcos and Gregor Stiglic and Allan Tucker",

booktitle = "Artificial Intelligence in Medicine - 21st International Conference on Artificial Intelligence in Medicine, AIME 2023, Proceedings",

address = "Germany",

}

Hansen, ER, Nielsen, TD, Mulvad, T, Strausholm, MN, Sagi, T & Hose, K 2023, Patient Event Sequences for Predicting Hospitalization Length of Stay. in JM Juarez, M Marcos, G Stiglic & A Tucker (eds), Artificial Intelligence in Medicine - 21st International Conference on Artificial Intelligence in Medicine, AIME 2023, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 13897 LNAI, Springer Science and Business Media Deutschland GmbH, pp. 51-56, 21st International Conference on Artificial Intelligence in Medicine, AIME 2023, Portoroz, Slovenia, 12/06/23. https://doi.org/10.1007/978-3-031-34344-5_7

Patient Event Sequences for Predicting Hospitalization Length of Stay. / Hansen, Emil Riis; Nielsen, Thomas Dyhre; Mulvad, Thomas et al.
Artificial Intelligence in Medicine - 21st International Conference on Artificial Intelligence in Medicine, AIME 2023, Proceedings. ed. / Jose M. Juarez; Mar Marcos; Gregor Stiglic; Allan Tucker. Springer Science and Business Media Deutschland GmbH, 2023. p. 51-56 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13897 LNAI).

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

TY - GEN

T1 - Patient Event Sequences for Predicting Hospitalization Length of Stay

AU - Hansen, Emil Riis

AU - Nielsen, Thomas Dyhre

AU - Mulvad, Thomas

AU - Strausholm, Mads Nibe

AU - Sagi, Tomer

AU - Hose, Katja

PY - 2023

Y1 - 2023

N2 - Predicting patients’ hospital length of stay (LOS) is essential for improving resource allocation and supporting decision-making in healthcare organizations. This paper proposes a novel transformer-based model, termed Medic-BERT (M-BERT), for predicting LOS by modeling patient information as sequences of events. We performed empirical experiments on a cohort of 48k emergency care patients from a large Danish hospital. Experimental results show that M-BERT can achieve high accuracy on a variety of LOS problems and outperforms traditional non-sequence-based machine learning approaches.

AB - Predicting patients’ hospital length of stay (LOS) is essential for improving resource allocation and supporting decision-making in healthcare organizations. This paper proposes a novel transformer-based model, termed Medic-BERT (M-BERT), for predicting LOS by modeling patient information as sequences of events. We performed empirical experiments on a cohort of 48k emergency care patients from a large Danish hospital. Experimental results show that M-BERT can achieve high accuracy on a variety of LOS problems and outperforms traditional non-sequence-based machine learning approaches.

KW - length of stay prediction

KW - sequence models

KW - transformers

UR - http://www.scopus.com/inward/record.url?scp=85151086836&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-34344-5_7

DO - 10.1007/978-3-031-34344-5_7

M3 - Conference contribution

AN - SCOPUS:85151086836

SN - 9783031343438

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 51

EP - 56

BT - Artificial Intelligence in Medicine - 21st International Conference on Artificial Intelligence in Medicine, AIME 2023, Proceedings

A2 - Juarez, Jose M.

A2 - Marcos, Mar

A2 - Stiglic, Gregor

A2 - Tucker, Allan

PB - Springer Science and Business Media Deutschland GmbH

T2 - 21st International Conference on Artificial Intelligence in Medicine, AIME 2023

Y2 - 12 June 2023 through 15 June 2023

ER -

Hansen ER, Nielsen TD, Mulvad T, Strausholm MN, Sagi T, Hose K. Patient Event Sequences for Predicting Hospitalization Length of Stay. In Juarez JM, Marcos M, Stiglic G, Tucker A, editors, Artificial Intelligence in Medicine - 21st International Conference on Artificial Intelligence in Medicine, AIME 2023, Proceedings. Springer Science and Business Media Deutschland GmbH. 2023. p. 51-56. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-34344-5_7