Data augmentation of wearable sensor data for Parkinson's disease monitoring using convolutional neural networks

Terry T. Um; Franz M.J. Pfister; Daniel Pichler; Satoshi Endo; Muriel Lang; Sandra Hirche; Urban Fietzek; Dana Kulic

doi:10.1145/3136755.3136817

Data augmentation of wearable sensor data for Parkinson's disease monitoring using convolutional neural networks

Terry T. Um, Franz M.J. Pfister, Daniel Pichler, Satoshi Endo, Muriel Lang, Sandra Hirche, Urban Fietzek, Dana Kulic

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

621 Citations (Scopus)

Abstract

While convolutional neural networks (CNNs) have been successfully applied to many challenging classification applications, they typically require large datasets for training. When the availability of labeled data is limited, data augmentation is a critical preprocessing step for CNNs. However, data augmentation for wearable sensor data has not been deeply investigated yet. In this paper, various data augmentation methods for wearable sensor data are proposed. The proposed methods and CNNs are applied to the classification of the motor state of Parkinson's Disease patients, which is challenging due to small dataset size, noisy labels, and large intra-class variability. Appropriate augmentation improves the classification performance from 77.54% to 86.88%.

Original language	English
Title of host publication	Proceedings of the 19th ACM International Conference on Multimodal Interaction
Editors	Edward Lank, Alessandro Vinciarelli, Eve Hoggan, Sriram Subramanian, Stephen A. Brewster
Place of Publication	New York NY USA
Publisher	Association for Computing Machinery (ACM)
Pages	216-220
Number of pages	5
ISBN (Electronic)	9781450355438
DOIs	https://doi.org/10.1145/3136755.3136817
Publication status	Published - 3 Nov 2017
Externally published	Yes
Event	International Conference on Multimodal Interfaces 2017 - Glasgow, United Kingdom Duration: 13 Nov 2017 → 17 Nov 2017 Conference number: 19th https://icmi.acm.org/2017/ (conference website) https://dl.acm.org/doi/proceedings/10.1145/3136755 (Proceedings)

Conference

Conference	International Conference on Multimodal Interfaces 2017
Abbreviated title	ICMI 2017
Country/Territory	United Kingdom
City	Glasgow
Period	13/11/17 → 17/11/17
Internet address	https://icmi.acm.org/2017/ (conference website) https://dl.acm.org/doi/proceedings/10.1145/3136755 (Proceedings)

Keywords

Convolutional neural networks
Data augmentation
Health monitoring
Parkinson's disease
Wearable sensor

UN SDGs

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

Access to Document

10.1145/3136755.3136817

Cite this

Um, T. T., Pfister, F. M. J., Pichler, D., Endo, S., Lang, M., Hirche, S., Fietzek, U., & Kulic, D. (2017). Data augmentation of wearable sensor data for Parkinson's disease monitoring using convolutional neural networks. In E. Lank, A. Vinciarelli, E. Hoggan, S. Subramanian, & S. A. Brewster (Eds.), Proceedings of the 19th ACM International Conference on Multimodal Interaction (pp. 216-220). Association for Computing Machinery (ACM). https://doi.org/10.1145/3136755.3136817

Um, Terry T. ; Pfister, Franz M.J. ; Pichler, Daniel et al. / Data augmentation of wearable sensor data for Parkinson's disease monitoring using convolutional neural networks. Proceedings of the 19th ACM International Conference on Multimodal Interaction. editor / Edward Lank ; Alessandro Vinciarelli ; Eve Hoggan ; Sriram Subramanian ; Stephen A. Brewster. New York NY USA : Association for Computing Machinery (ACM), 2017. pp. 216-220

@inproceedings{81c04c79d6c245c59b0d76af44a39054,

title = "Data augmentation of wearable sensor data for Parkinson's disease monitoring using convolutional neural networks",

abstract = "While convolutional neural networks (CNNs) have been successfully applied to many challenging classification applications, they typically require large datasets for training. When the availability of labeled data is limited, data augmentation is a critical preprocessing step for CNNs. However, data augmentation for wearable sensor data has not been deeply investigated yet. In this paper, various data augmentation methods for wearable sensor data are proposed. The proposed methods and CNNs are applied to the classification of the motor state of Parkinson's Disease patients, which is challenging due to small dataset size, noisy labels, and large intra-class variability. Appropriate augmentation improves the classification performance from 77.54% to 86.88%.",

keywords = "Convolutional neural networks, Data augmentation, Health monitoring, Parkinson's disease, Wearable sensor",

author = "Um, {Terry T.} and Pfister, {Franz M.J.} and Daniel Pichler and Satoshi Endo and Muriel Lang and Sandra Hirche and Urban Fietzek and Dana Kulic",

year = "2017",

month = nov,

day = "3",

doi = "10.1145/3136755.3136817",

language = "English",

pages = "216--220",

editor = "Edward Lank and Alessandro Vinciarelli and Eve Hoggan and Sriram Subramanian and Brewster, {Stephen A.}",

booktitle = "Proceedings of the 19th ACM International Conference on Multimodal Interaction",

publisher = "Association for Computing Machinery (ACM)",

address = "United States of America",

note = "International Conference on Multimodal Interfaces 2017, ICMI 2017 ; Conference date: 13-11-2017 Through 17-11-2017",

url = "https://icmi.acm.org/2017/, https://dl.acm.org/doi/proceedings/10.1145/3136755",

}

Um, TT, Pfister, FMJ, Pichler, D, Endo, S, Lang, M, Hirche, S, Fietzek, U & Kulic, D 2017, Data augmentation of wearable sensor data for Parkinson's disease monitoring using convolutional neural networks. in E Lank, A Vinciarelli, E Hoggan, S Subramanian & SA Brewster (eds), Proceedings of the 19th ACM International Conference on Multimodal Interaction. Association for Computing Machinery (ACM), New York NY USA, pp. 216-220, International Conference on Multimodal Interfaces 2017, Glasgow, United Kingdom, 13/11/17. https://doi.org/10.1145/3136755.3136817

Data augmentation of wearable sensor data for Parkinson's disease monitoring using convolutional neural networks. / Um, Terry T.; Pfister, Franz M.J.; Pichler, Daniel et al.
Proceedings of the 19th ACM International Conference on Multimodal Interaction. ed. / Edward Lank; Alessandro Vinciarelli; Eve Hoggan; Sriram Subramanian; Stephen A. Brewster. New York NY USA: Association for Computing Machinery (ACM), 2017. p. 216-220.

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

TY - GEN

T1 - Data augmentation of wearable sensor data for Parkinson's disease monitoring using convolutional neural networks

AU - Um, Terry T.

AU - Pfister, Franz M.J.

AU - Pichler, Daniel

AU - Endo, Satoshi

AU - Lang, Muriel

AU - Hirche, Sandra

AU - Fietzek, Urban

AU - Kulic, Dana

N1 - Conference code: 19th

PY - 2017/11/3

Y1 - 2017/11/3

N2 - While convolutional neural networks (CNNs) have been successfully applied to many challenging classification applications, they typically require large datasets for training. When the availability of labeled data is limited, data augmentation is a critical preprocessing step for CNNs. However, data augmentation for wearable sensor data has not been deeply investigated yet. In this paper, various data augmentation methods for wearable sensor data are proposed. The proposed methods and CNNs are applied to the classification of the motor state of Parkinson's Disease patients, which is challenging due to small dataset size, noisy labels, and large intra-class variability. Appropriate augmentation improves the classification performance from 77.54% to 86.88%.

AB - While convolutional neural networks (CNNs) have been successfully applied to many challenging classification applications, they typically require large datasets for training. When the availability of labeled data is limited, data augmentation is a critical preprocessing step for CNNs. However, data augmentation for wearable sensor data has not been deeply investigated yet. In this paper, various data augmentation methods for wearable sensor data are proposed. The proposed methods and CNNs are applied to the classification of the motor state of Parkinson's Disease patients, which is challenging due to small dataset size, noisy labels, and large intra-class variability. Appropriate augmentation improves the classification performance from 77.54% to 86.88%.

KW - Convolutional neural networks

KW - Data augmentation

KW - Health monitoring

KW - Parkinson's disease

KW - Wearable sensor

UR - http://www.scopus.com/inward/record.url?scp=85046336429&partnerID=8YFLogxK

U2 - 10.1145/3136755.3136817

DO - 10.1145/3136755.3136817

M3 - Conference Paper

AN - SCOPUS:85046336429

SP - 216

EP - 220

BT - Proceedings of the 19th ACM International Conference on Multimodal Interaction

A2 - Lank, Edward

A2 - Vinciarelli, Alessandro

A2 - Hoggan, Eve

A2 - Subramanian, Sriram

A2 - Brewster, Stephen A.

PB - Association for Computing Machinery (ACM)

CY - New York NY USA

T2 - International Conference on Multimodal Interfaces 2017

Y2 - 13 November 2017 through 17 November 2017

ER -

Um TT, Pfister FMJ, Pichler D, Endo S, Lang M, Hirche S et al. Data augmentation of wearable sensor data for Parkinson's disease monitoring using convolutional neural networks. In Lank E, Vinciarelli A, Hoggan E, Subramanian S, Brewster SA, editors, Proceedings of the 19th ACM International Conference on Multimodal Interaction. New York NY USA: Association for Computing Machinery (ACM). 2017. p. 216-220 doi: 10.1145/3136755.3136817

Data augmentation of wearable sensor data for Parkinson's disease monitoring using convolutional neural networks

Abstract

Conference

Keywords

UN SDGs

Access to Document

Other files and links

Cite this