Classification of implantable rotary blood pump states with class noise

Hui Lee Ooi; Manjeevan Seera; Siew Cheok Ng; Chee Peng Lim; Chu Kiong Loo; Nigel H. Lovell; Stephen J. Redmond; Einly Lim

doi:10.1109/JBHI.2015.2412375

Classification of implantable rotary blood pump states with class noise

Hui Lee Ooi, Manjeevan Seera, Siew Cheok Ng, Chee Peng Lim, Chu Kiong Loo, Nigel H. Lovell, Stephen J. Redmond, Einly Lim

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

Abstract

A medical case study related to implantable rotary blood pumps is examined. Five classifiers and two ensemble classifiers are applied to process the signals collected from the pumps for the identification of the aortic valve nonopening pump state. In addition to the noise-free datasets, up to 40\% class noise has been added to the signals to evaluate the classification performance when mislabeling is present in the classifier training set. In order to ensure a reliable diagnostic model for the identification of the pump states, classifications performed with and without class noise are evaluated. The multilayer perceptron emerged as the best performing classifier for pump state detection due to its high accuracy as well as robustness against class noise.

Original language	English
Pages (from-to)	829-837
Number of pages	9
Journal	IEEE Journal of Biomedical and Health Informatics
Volume	20
Issue number	3
DOIs	https://doi.org/10.1109/JBHI.2015.2412375
Publication status	Published - May 2016
Externally published	Yes

Keywords

class noise
classification
classifier
ensemble classifier
implantable rotary blood pump
left ventricular assist device
mislabeling
pump state classification

Access to Document

10.1109/JBHI.2015.2412375

Cite this

@article{575adb03478c4daa97f40c0db6ce96bb,

title = "Classification of implantable rotary blood pump states with class noise",

abstract = "A medical case study related to implantable rotary blood pumps is examined. Five classifiers and two ensemble classifiers are applied to process the signals collected from the pumps for the identification of the aortic valve nonopening pump state. In addition to the noise-free datasets, up to 40\% class noise has been added to the signals to evaluate the classification performance when mislabeling is present in the classifier training set. In order to ensure a reliable diagnostic model for the identification of the pump states, classifications performed with and without class noise are evaluated. The multilayer perceptron emerged as the best performing classifier for pump state detection due to its high accuracy as well as robustness against class noise.",

keywords = "class noise, classification, classifier, ensemble classifier, implantable rotary blood pump, left ventricular assist device, mislabeling, pump state classification",

author = "Ooi, {Hui Lee} and Manjeevan Seera and Ng, {Siew Cheok} and Lim, {Chee Peng} and Loo, {Chu Kiong} and Lovell, {Nigel H.} and Redmond, {Stephen J.} and Einly Lim",

note = "Funding Information: This work was supported partially by Ministry of Higher Education Malaysia (Grant number: UM/HIR (MOHE)/ENG/50), UMRG Research Subprogram (Project Number RP003D-13ICT), and the Australian Research Council Linkages scheme Publisher Copyright: {\textcopyright} 2013 IEEE. Copyright: Copyright 2016 Elsevier B.V., All rights reserved.",

year = "2016",

month = may,

doi = "10.1109/JBHI.2015.2412375",

language = "English",

volume = "20",

pages = "829--837",

journal = "IEEE Journal of Biomedical and Health Informatics",

issn = "2168-2194",

publisher = "IEEE, Institute of Electrical and Electronics Engineers",

number = "3",

}

TY - JOUR

T1 - Classification of implantable rotary blood pump states with class noise

AU - Ooi, Hui Lee

AU - Seera, Manjeevan

AU - Ng, Siew Cheok

AU - Lim, Chee Peng

AU - Loo, Chu Kiong

AU - Lovell, Nigel H.

AU - Redmond, Stephen J.

AU - Lim, Einly

N1 - Funding Information: This work was supported partially by Ministry of Higher Education Malaysia (Grant number: UM/HIR (MOHE)/ENG/50), UMRG Research Subprogram (Project Number RP003D-13ICT), and the Australian Research Council Linkages scheme Publisher Copyright: © 2013 IEEE. Copyright: Copyright 2016 Elsevier B.V., All rights reserved.

PY - 2016/5

Y1 - 2016/5

N2 - A medical case study related to implantable rotary blood pumps is examined. Five classifiers and two ensemble classifiers are applied to process the signals collected from the pumps for the identification of the aortic valve nonopening pump state. In addition to the noise-free datasets, up to 40\% class noise has been added to the signals to evaluate the classification performance when mislabeling is present in the classifier training set. In order to ensure a reliable diagnostic model for the identification of the pump states, classifications performed with and without class noise are evaluated. The multilayer perceptron emerged as the best performing classifier for pump state detection due to its high accuracy as well as robustness against class noise.

AB - A medical case study related to implantable rotary blood pumps is examined. Five classifiers and two ensemble classifiers are applied to process the signals collected from the pumps for the identification of the aortic valve nonopening pump state. In addition to the noise-free datasets, up to 40\% class noise has been added to the signals to evaluate the classification performance when mislabeling is present in the classifier training set. In order to ensure a reliable diagnostic model for the identification of the pump states, classifications performed with and without class noise are evaluated. The multilayer perceptron emerged as the best performing classifier for pump state detection due to its high accuracy as well as robustness against class noise.

KW - class noise

KW - classification

KW - classifier

KW - ensemble classifier

KW - implantable rotary blood pump

KW - left ventricular assist device

KW - mislabeling

KW - pump state classification

UR - http://www.scopus.com/inward/record.url?scp=84969921286&partnerID=8YFLogxK

U2 - 10.1109/JBHI.2015.2412375

DO - 10.1109/JBHI.2015.2412375

M3 - Article

C2 - 25781963

AN - SCOPUS:84969921286

VL - 20

SP - 829

EP - 837

JO - IEEE Journal of Biomedical and Health Informatics

JF - IEEE Journal of Biomedical and Health Informatics

SN - 2168-2194

IS - 3

ER -

Classification of implantable rotary blood pump states with class noise

Abstract

Keywords

Access to Document

Other files and links

Cite this