Multitask Recalibrated Aggregation Network for Medical Code Prediction

Wei Sun; Shaoxiong Ji; Erik Cambria; Pekka Marttinen

doi:10.1007/978-3-030-86514-6_23

Multitask Recalibrated Aggregation Network for Medical Code Prediction

Wei Sun
, Shaoxiong Ji^*
, Erik Cambria
, Pekka Marttinen

^*Corresponding author for this work

Aalto University
Nanyang Technological University

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

10 Citations (Scopus)

69 Downloads (Pure)

Abstract

Medical coding translates professionally written medical reports into standardized codes, which is an essential part of medical information systems and health insurance reimbursement. Manual coding by trained human coders is time-consuming and error-prone. Thus, automated coding algorithms have been developed, building especially on the recent advances in machine learning and deep neural networks. To solve the challenges of encoding lengthy and noisy clinical documents and capturing code associations, we propose a multitask recalibrated aggregation network. In particular, multitask learning shares information across different coding schemes and captures the dependencies between different medical codes. Feature recalibration and aggregation in shared modules enhance representation learning for lengthy notes. Experiments with a real-world MIMIC-III dataset show significantly improved predictive performance.

Original language	English
Title of host publication	Machine Learning and Knowledge Discovery in Databases
Subtitle of host publication	Applied Data Science Track - European Conference, ECML PKDD 2021, Proceedings
Editors	Yuxiao Dong, Nicolas Kourtellis, Barbara Hammer, Jose A. Lozano
Publisher	Springer
Pages	367-383
Number of pages	17
ISBN (Print)	978-3-030-86513-9
DOIs	https://doi.org/10.1007/978-3-030-86514-6_23
Publication status	Published - 2021
MoE publication type	A4 Conference publication
Event	European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases - Virtual, Online Duration: 13 Sept 2021 → 17 Sept 2021

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Publisher	Springer
Volume	12978 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases
Abbreviated title	ECML PKDD
City	Virtual, Online
Period	13/09/2021 → 17/09/2021

Funding

Acknowledgments. This work was supported by the Academy of Finland (grant 336033) and EU H2020 (grant 101016775). We acknowledge the computational resources provided by the Aalto Science-IT project. The authors wish to acknowledge CSC - IT Center for Science, Finland, for computational resources.

Keywords

Medical code prediction
Multitask learning
Recalibrated aggregation network

Access to Document

10.1007/978-3-030-86514-6_23Licence: CC BY

Sun_Multitask Recalibrated Aggregation Network for Medical Code PredictionFinal published version, 479 KBLicence: CC BY

Cite this

Sun, W., Ji, S., Cambria, E., & Marttinen, P. (2021). Multitask Recalibrated Aggregation Network for Medical Code Prediction. In Y. Dong, N. Kourtellis, B. Hammer, & J. A. Lozano (Eds.), Machine Learning and Knowledge Discovery in Databases: Applied Data Science Track - European Conference, ECML PKDD 2021, Proceedings (pp. 367-383). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12978 LNAI). Springer. https://doi.org/10.1007/978-3-030-86514-6_23

Sun, Wei ; Ji, Shaoxiong ; Cambria, Erik et al. / Multitask Recalibrated Aggregation Network for Medical Code Prediction. Machine Learning and Knowledge Discovery in Databases: Applied Data Science Track - European Conference, ECML PKDD 2021, Proceedings. editor / Yuxiao Dong ; Nicolas Kourtellis ; Barbara Hammer ; Jose A. Lozano. Springer, 2021. pp. 367-383 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{c85cb8f0432d455ba57f2ec7f5477684,

title = "Multitask Recalibrated Aggregation Network for Medical Code Prediction",

abstract = "Medical coding translates professionally written medical reports into standardized codes, which is an essential part of medical information systems and health insurance reimbursement. Manual coding by trained human coders is time-consuming and error-prone. Thus, automated coding algorithms have been developed, building especially on the recent advances in machine learning and deep neural networks. To solve the challenges of encoding lengthy and noisy clinical documents and capturing code associations, we propose a multitask recalibrated aggregation network. In particular, multitask learning shares information across different coding schemes and captures the dependencies between different medical codes. Feature recalibration and aggregation in shared modules enhance representation learning for lengthy notes. Experiments with a real-world MIMIC-III dataset show significantly improved predictive performance.",

keywords = "Medical code prediction, Multitask learning, Recalibrated aggregation network",

author = "Wei Sun and Shaoxiong Ji and Erik Cambria and Pekka Marttinen",

note = "| openaire: EC/H2020/101016775/EU//INTERVENE Funding Information: Acknowledgments. This work was supported by the Academy of Finland (grant 336033) and EU H2020 (grant 101016775). We acknowledge the computational resources provided by the Aalto Science-IT project. The authors wish to acknowledge CSC - IT Center for Science, Finland, for computational resources. Publisher Copyright: {\textcopyright} 2021, The Author(s).; European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, ECML PKDD ; Conference date: 13-09-2021 Through 17-09-2021",

year = "2021",

doi = "10.1007/978-3-030-86514-6\_23",

language = "English",

isbn = "978-3-030-86513-9",

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

publisher = "Springer",

pages = "367--383",

editor = "Yuxiao Dong and Nicolas Kourtellis and Barbara Hammer and Lozano, \{Jose A.\}",

booktitle = "Machine Learning and Knowledge Discovery in Databases",

address = "Germany",

}

Sun, W, Ji, S, Cambria, E & Marttinen, P 2021, Multitask Recalibrated Aggregation Network for Medical Code Prediction. in Y Dong, N Kourtellis, B Hammer & JA Lozano (eds), Machine Learning and Knowledge Discovery in Databases: Applied Data Science Track - European Conference, ECML PKDD 2021, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 12978 LNAI, Springer, pp. 367-383, European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, Virtual, Online, 13/09/2021. https://doi.org/10.1007/978-3-030-86514-6_23

Multitask Recalibrated Aggregation Network for Medical Code Prediction. / Sun, Wei; Ji, Shaoxiong; Cambria, Erik et al.
Machine Learning and Knowledge Discovery in Databases: Applied Data Science Track - European Conference, ECML PKDD 2021, Proceedings. ed. / Yuxiao Dong; Nicolas Kourtellis; Barbara Hammer; Jose A. Lozano. Springer, 2021. p. 367-383 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12978 LNAI).

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

TY - GEN

T1 - Multitask Recalibrated Aggregation Network for Medical Code Prediction

AU - Sun, Wei

AU - Ji, Shaoxiong

AU - Cambria, Erik

AU - Marttinen, Pekka

N1 - | openaire: EC/H2020/101016775/EU//INTERVENE Funding Information: Acknowledgments. This work was supported by the Academy of Finland (grant 336033) and EU H2020 (grant 101016775). We acknowledge the computational resources provided by the Aalto Science-IT project. The authors wish to acknowledge CSC - IT Center for Science, Finland, for computational resources. Publisher Copyright: © 2021, The Author(s).

PY - 2021

Y1 - 2021

N2 - Medical coding translates professionally written medical reports into standardized codes, which is an essential part of medical information systems and health insurance reimbursement. Manual coding by trained human coders is time-consuming and error-prone. Thus, automated coding algorithms have been developed, building especially on the recent advances in machine learning and deep neural networks. To solve the challenges of encoding lengthy and noisy clinical documents and capturing code associations, we propose a multitask recalibrated aggregation network. In particular, multitask learning shares information across different coding schemes and captures the dependencies between different medical codes. Feature recalibration and aggregation in shared modules enhance representation learning for lengthy notes. Experiments with a real-world MIMIC-III dataset show significantly improved predictive performance.

AB - Medical coding translates professionally written medical reports into standardized codes, which is an essential part of medical information systems and health insurance reimbursement. Manual coding by trained human coders is time-consuming and error-prone. Thus, automated coding algorithms have been developed, building especially on the recent advances in machine learning and deep neural networks. To solve the challenges of encoding lengthy and noisy clinical documents and capturing code associations, we propose a multitask recalibrated aggregation network. In particular, multitask learning shares information across different coding schemes and captures the dependencies between different medical codes. Feature recalibration and aggregation in shared modules enhance representation learning for lengthy notes. Experiments with a real-world MIMIC-III dataset show significantly improved predictive performance.

KW - Medical code prediction

KW - Multitask learning

KW - Recalibrated aggregation network

UR - https://www.scopus.com/pages/publications/85115712795

U2 - 10.1007/978-3-030-86514-6_23

DO - 10.1007/978-3-030-86514-6_23

M3 - Conference article in proceedings

AN - SCOPUS:85115712795

SN - 978-3-030-86513-9

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

SP - 367

EP - 383

BT - Machine Learning and Knowledge Discovery in Databases

A2 - Dong, Yuxiao

A2 - Kourtellis, Nicolas

A2 - Hammer, Barbara

A2 - Lozano, Jose A.

PB - Springer

T2 - European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases

Y2 - 13 September 2021 through 17 September 2021

ER -

Sun W, Ji S, Cambria E, Marttinen P. Multitask Recalibrated Aggregation Network for Medical Code Prediction. In Dong Y, Kourtellis N, Hammer B, Lozano JA, editors, Machine Learning and Knowledge Discovery in Databases: Applied Data Science Track - European Conference, ECML PKDD 2021, Proceedings. Springer. 2021. p. 367-383. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-86514-6_23

Multitask Recalibrated Aggregation Network for Medical Code Prediction

Abstract

Publication series

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

INTERVENE: International consortium for integrative genomics prediction

DATALIT: Data Literacy for Responsible Decision-Making

Science-IT

Cite this

Multitask Recalibrated Aggregation Network for Medical Code Prediction

Abstract

Publication series

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Projects

INTERVENE: International consortium for integrative genomics prediction

DATALIT: Data Literacy for Responsible Decision-Making

Equipment

Science-IT

Cite this