Multi-task deep neural networks for automated extraction of primary site and laterality information from cancer pathology reports

Hong Jun Yoon; Arvind Ramanathan; Georgia Tourassi

doi:10.1007/978-3-319-47898-2_21

Multi-task deep neural networks for automated extraction of primary site and laterality information from cancer pathology reports

Hong Jun Yoon, Arvind Ramanathan, Georgia Tourassi

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

18 Scopus citations

Abstract

Automated annotation of free-text cancer pathology reports is a critical challenge for cancer registries and the national cancer surveillance program. In this paper, we investigated deep neural networks (DNNs) for automated extraction of the primary cancer site and its laterality, two fundamental targets of cancer reporting. Our experiments showed that single-task DNNs are capable of extracting information with higher precision and recall than traditional classification methods for the more challenging target. Furthermore, a multi-task learning DNN resulted in further performance improvement. This preliminary study, indicate the strong potential for multi-task deep neural networks to extract cancer-relevant information from free-text pathology reports.

Original language	English
Title of host publication	Advances in Big Data - Proceedings of the 2nd INNS Conference on Big Data, 2016
Editors	Asim Roy, Marley Vellasco, Yannis Manolopoulos, Lazaros Iliadis, Plamen Angelov
Publisher	Springer Verlag
Pages	195-204
Number of pages	10
ISBN (Print)	9783319478975
DOIs	https://doi.org/10.1007/978-3-319-47898-2_21
State	Published - 2017
Event	2nd International Neural Network Society Conference on Big Data, INNS 2016 - Thessaloniki, Greece Duration: Oct 23 2016 → Oct 25 2016

Publication series

Name	Advances in Intelligent Systems and Computing
Volume	529
ISSN (Print)	2194-5357

Conference

Conference	2nd International Neural Network Society Conference on Big Data, INNS 2016
Country/Territory	Greece
City	Thessaloniki
Period	10/23/16 → 10/25/16

Funding

This manuscript has been authored by UT-Battelle, LLC under Contract No. DE-AC05-00OR22725 with the U.S. Department of Energy. The United States Government retains and the publisher, by accepting the article for publication, acknowledges that the United States Government retains a non-exclusive, paid-up, irrevocable, world-wide license to publish or reproduce the published form of this manuscript, or allow others to do so, for United States Government purposes. The Department of Energy will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan ( http://energy.gov/downloads/doe-public-access-plan ). The study was supported by the Laboratory Directed Research and Development (LDRD) program of Oak Ridge National Laboratory, under LDRD projects No. 7417 and No. 8231.

Funders	Funder number
U.S. Department of Energy
Laboratory Directed Research and Development	8231, 7417

Keywords

Cancer pathology report
Deep neural network
Multi-task learning
Natural language processing

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1007/978-3-319-47898-2_21

Cite this

Yoon, H. J., Ramanathan, A., & Tourassi, G. (2017). Multi-task deep neural networks for automated extraction of primary site and laterality information from cancer pathology reports. In A. Roy, M. Vellasco, Y. Manolopoulos, L. Iliadis, & P. Angelov (Eds.), Advances in Big Data - Proceedings of the 2nd INNS Conference on Big Data, 2016 (pp. 195-204). (Advances in Intelligent Systems and Computing; Vol. 529). Springer Verlag. https://doi.org/10.1007/978-3-319-47898-2_21

Yoon, Hong Jun ; Ramanathan, Arvind ; Tourassi, Georgia. / Multi-task deep neural networks for automated extraction of primary site and laterality information from cancer pathology reports. Advances in Big Data - Proceedings of the 2nd INNS Conference on Big Data, 2016. editor / Asim Roy ; Marley Vellasco ; Yannis Manolopoulos ; Lazaros Iliadis ; Plamen Angelov. Springer Verlag, 2017. pp. 195-204 (Advances in Intelligent Systems and Computing).

@inproceedings{73c466a350e1463eae0acbd540bbedcf,

title = "Multi-task deep neural networks for automated extraction of primary site and laterality information from cancer pathology reports",

abstract = "Automated annotation of free-text cancer pathology reports is a critical challenge for cancer registries and the national cancer surveillance program. In this paper, we investigated deep neural networks (DNNs) for automated extraction of the primary cancer site and its laterality, two fundamental targets of cancer reporting. Our experiments showed that single-task DNNs are capable of extracting information with higher precision and recall than traditional classification methods for the more challenging target. Furthermore, a multi-task learning DNN resulted in further performance improvement. This preliminary study, indicate the strong potential for multi-task deep neural networks to extract cancer-relevant information from free-text pathology reports.",

keywords = "Cancer pathology report, Deep neural network, Multi-task learning, Natural language processing",

author = "Yoon, {Hong Jun} and Arvind Ramanathan and Georgia Tourassi",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing AG 2017.; 2nd International Neural Network Society Conference on Big Data, INNS 2016 ; Conference date: 23-10-2016 Through 25-10-2016",

year = "2017",

doi = "10.1007/978-3-319-47898-2_21",

language = "English",

isbn = "9783319478975",

series = "Advances in Intelligent Systems and Computing",

publisher = "Springer Verlag",

pages = "195--204",

editor = "Asim Roy and Marley Vellasco and Yannis Manolopoulos and Lazaros Iliadis and Plamen Angelov",

booktitle = "Advances in Big Data - Proceedings of the 2nd INNS Conference on Big Data, 2016",

}

Yoon, HJ, Ramanathan, A & Tourassi, G 2017, Multi-task deep neural networks for automated extraction of primary site and laterality information from cancer pathology reports. in A Roy, M Vellasco, Y Manolopoulos, L Iliadis & P Angelov (eds), Advances in Big Data - Proceedings of the 2nd INNS Conference on Big Data, 2016. Advances in Intelligent Systems and Computing, vol. 529, Springer Verlag, pp. 195-204, 2nd International Neural Network Society Conference on Big Data, INNS 2016, Thessaloniki, Greece, 10/23/16. https://doi.org/10.1007/978-3-319-47898-2_21

Multi-task deep neural networks for automated extraction of primary site and laterality information from cancer pathology reports. / Yoon, Hong Jun; Ramanathan, Arvind; Tourassi, Georgia.
Advances in Big Data - Proceedings of the 2nd INNS Conference on Big Data, 2016. ed. / Asim Roy; Marley Vellasco; Yannis Manolopoulos; Lazaros Iliadis; Plamen Angelov. Springer Verlag, 2017. p. 195-204 (Advances in Intelligent Systems and Computing; Vol. 529).

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

TY - GEN

T1 - Multi-task deep neural networks for automated extraction of primary site and laterality information from cancer pathology reports

AU - Yoon, Hong Jun

AU - Ramanathan, Arvind

AU - Tourassi, Georgia

N1 - Publisher Copyright: © Springer International Publishing AG 2017.

PY - 2017

Y1 - 2017

N2 - Automated annotation of free-text cancer pathology reports is a critical challenge for cancer registries and the national cancer surveillance program. In this paper, we investigated deep neural networks (DNNs) for automated extraction of the primary cancer site and its laterality, two fundamental targets of cancer reporting. Our experiments showed that single-task DNNs are capable of extracting information with higher precision and recall than traditional classification methods for the more challenging target. Furthermore, a multi-task learning DNN resulted in further performance improvement. This preliminary study, indicate the strong potential for multi-task deep neural networks to extract cancer-relevant information from free-text pathology reports.

AB - Automated annotation of free-text cancer pathology reports is a critical challenge for cancer registries and the national cancer surveillance program. In this paper, we investigated deep neural networks (DNNs) for automated extraction of the primary cancer site and its laterality, two fundamental targets of cancer reporting. Our experiments showed that single-task DNNs are capable of extracting information with higher precision and recall than traditional classification methods for the more challenging target. Furthermore, a multi-task learning DNN resulted in further performance improvement. This preliminary study, indicate the strong potential for multi-task deep neural networks to extract cancer-relevant information from free-text pathology reports.

KW - Cancer pathology report

KW - Deep neural network

KW - Multi-task learning

KW - Natural language processing

UR - http://www.scopus.com/inward/record.url?scp=84994525473&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-47898-2_21

DO - 10.1007/978-3-319-47898-2_21

M3 - Conference contribution

AN - SCOPUS:84994525473

SN - 9783319478975

T3 - Advances in Intelligent Systems and Computing

SP - 195

EP - 204

BT - Advances in Big Data - Proceedings of the 2nd INNS Conference on Big Data, 2016

A2 - Roy, Asim

A2 - Vellasco, Marley

A2 - Manolopoulos, Yannis

A2 - Iliadis, Lazaros

A2 - Angelov, Plamen

PB - Springer Verlag

T2 - 2nd International Neural Network Society Conference on Big Data, INNS 2016

Y2 - 23 October 2016 through 25 October 2016

ER -

Yoon HJ, Ramanathan A, Tourassi G. Multi-task deep neural networks for automated extraction of primary site and laterality information from cancer pathology reports. In Roy A, Vellasco M, Manolopoulos Y, Iliadis L, Angelov P, editors, Advances in Big Data - Proceedings of the 2nd INNS Conference on Big Data, 2016. Springer Verlag. 2017. p. 195-204. (Advances in Intelligent Systems and Computing). doi: 10.1007/978-3-319-47898-2_21

Multi-task deep neural networks for automated extraction of primary site and laterality information from cancer pathology reports

Abstract

Publication series

Conference

Funding

Keywords

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this