Using multimodal biosignal data from wearables to detect focal motor seizures in individual epilepsy patients

Sebastian Böttcher; Nikolay V. Manyakov; Nino Epitashvili; Amos Folarin; Mark P. Richardson; Matthias Dümpelmann; Andreas Schulze-Bonhage; Kristof van Laerhoven

doi:10.1145/3361684.3361687

Using multimodal biosignal data from wearables to detect focal motor seizures in individual epilepsy patients

Sebastian Böttcher^*, Nikolay V. Manyakov, Nino Epitashvili, Amos Folarin, Mark P. Richardson, Matthias Dümpelmann, Andreas Schulze-Bonhage, Kristof van Laerhoven

^*Corresponding author for this work

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

5 Citations (Scopus)

Abstract

Epilepsy seizure detection with wearable devices is an emerging research field. As opposed to the gold standard, consisting of simultaneous video and EEG monitoring of patients, wearables have the advantage that they put a lower burden on epilepsy patients. We report on the first stages in a research effort that is dedicated to the development of a multimodal seizure detection system specifically for focal onset epileptic seizures. By in-depth analysis of data from three in-hospital patients with each having six to nine seizures recorded, we show that such seizures can manifest very differently and thus significantly impact classification. Using a Random Forest model on a rich set of features, we have obtained overall precision and recall scores of up to 0.92 and 0.72 respectively. These results show that the approach has validity, but we identify the type of focal seizure to be a critical factor for the classification performance.

Original language	English
Title of host publication	iWOAR 2019 - 6th International Workshop on Sensor-Based Activity Recognition and Interaction, Proceedings
Editors	Stefan Ludtke, Sebastian Bader, Kristina Yordanova, Thomas Kirste
Publisher	Association for Computing Machinery
ISBN (Electronic)	9781450377140
DOIs	https://doi.org/10.1145/3361684.3361687
Publication status	Published - 16 Sept 2019
Event	6th International Workshop on Sensor-Based Activity Recognition and Interaction, iWOAR 2019 - Rostock, Germany Duration: 16 Sept 2019 → 17 Sept 2019

Conference

Conference	6th International Workshop on Sensor-Based Activity Recognition and Interaction, iWOAR 2019
Country/Territory	Germany
City	Rostock
Period	16/09/2019 → 17/09/2019

Keywords

Epilepsy seizure detection
Multimodal biosignals
Wearables

Access to Document

10.1145/3361684.3361687

Cite this

Böttcher, S., Manyakov, N. V., Epitashvili, N., Folarin, A., Richardson, M. P., Dümpelmann, M., Schulze-Bonhage, A., & van Laerhoven, K. (2019). Using multimodal biosignal data from wearables to detect focal motor seizures in individual epilepsy patients. In S. Ludtke, S. Bader, K. Yordanova, & T. Kirste (Eds.), iWOAR 2019 - 6th International Workshop on Sensor-Based Activity Recognition and Interaction, Proceedings Article 3361687 Association for Computing Machinery. https://doi.org/10.1145/3361684.3361687

Böttcher, Sebastian ; Manyakov, Nikolay V. ; Epitashvili, Nino et al. / Using multimodal biosignal data from wearables to detect focal motor seizures in individual epilepsy patients. iWOAR 2019 - 6th International Workshop on Sensor-Based Activity Recognition and Interaction, Proceedings. editor / Stefan Ludtke ; Sebastian Bader ; Kristina Yordanova ; Thomas Kirste. Association for Computing Machinery, 2019.

@inbook{eed3aa8de9a443c48171e6c7849c2bcb,

title = "Using multimodal biosignal data from wearables to detect focal motor seizures in individual epilepsy patients",

abstract = "Epilepsy seizure detection with wearable devices is an emerging research field. As opposed to the gold standard, consisting of simultaneous video and EEG monitoring of patients, wearables have the advantage that they put a lower burden on epilepsy patients. We report on the first stages in a research effort that is dedicated to the development of a multimodal seizure detection system specifically for focal onset epileptic seizures. By in-depth analysis of data from three in-hospital patients with each having six to nine seizures recorded, we show that such seizures can manifest very differently and thus significantly impact classification. Using a Random Forest model on a rich set of features, we have obtained overall precision and recall scores of up to 0.92 and 0.72 respectively. These results show that the approach has validity, but we identify the type of focal seizure to be a critical factor for the classification performance.",

keywords = "Epilepsy seizure detection, Multimodal biosignals, Wearables",

author = "Sebastian B{\"o}ttcher and Manyakov, {Nikolay V.} and Nino Epitashvili and Amos Folarin and Richardson, {Mark P.} and Matthias D{\"u}mpelmann and Andreas Schulze-Bonhage and {van Laerhoven}, Kristof",

year = "2019",

month = sep,

day = "16",

doi = "10.1145/3361684.3361687",

language = "English",

editor = "Stefan Ludtke and Sebastian Bader and Kristina Yordanova and Thomas Kirste",

booktitle = "iWOAR 2019 - 6th International Workshop on Sensor-Based Activity Recognition and Interaction, Proceedings",

publisher = "Association for Computing Machinery",

note = "6th International Workshop on Sensor-Based Activity Recognition and Interaction, iWOAR 2019 ; Conference date: 16-09-2019 Through 17-09-2019",

}

Böttcher, S, Manyakov, NV, Epitashvili, N, Folarin, A , Richardson, MP, Dümpelmann, M, Schulze-Bonhage, A & van Laerhoven, K 2019, Using multimodal biosignal data from wearables to detect focal motor seizures in individual epilepsy patients. in S Ludtke, S Bader, K Yordanova & T Kirste (eds), iWOAR 2019 - 6th International Workshop on Sensor-Based Activity Recognition and Interaction, Proceedings., 3361687, Association for Computing Machinery, 6th International Workshop on Sensor-Based Activity Recognition and Interaction, iWOAR 2019, Rostock, Germany, 16/09/2019. https://doi.org/10.1145/3361684.3361687

Using multimodal biosignal data from wearables to detect focal motor seizures in individual epilepsy patients. / Böttcher, Sebastian; Manyakov, Nikolay V.; Epitashvili, Nino et al.
iWOAR 2019 - 6th International Workshop on Sensor-Based Activity Recognition and Interaction, Proceedings. ed. / Stefan Ludtke; Sebastian Bader; Kristina Yordanova; Thomas Kirste. Association for Computing Machinery, 2019. 3361687.

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

TY - CHAP

T1 - Using multimodal biosignal data from wearables to detect focal motor seizures in individual epilepsy patients

AU - Böttcher, Sebastian

AU - Manyakov, Nikolay V.

AU - Epitashvili, Nino

AU - Folarin, Amos

AU - Richardson, Mark P.

AU - Dümpelmann, Matthias

AU - Schulze-Bonhage, Andreas

AU - van Laerhoven, Kristof

PY - 2019/9/16

Y1 - 2019/9/16

N2 - Epilepsy seizure detection with wearable devices is an emerging research field. As opposed to the gold standard, consisting of simultaneous video and EEG monitoring of patients, wearables have the advantage that they put a lower burden on epilepsy patients. We report on the first stages in a research effort that is dedicated to the development of a multimodal seizure detection system specifically for focal onset epileptic seizures. By in-depth analysis of data from three in-hospital patients with each having six to nine seizures recorded, we show that such seizures can manifest very differently and thus significantly impact classification. Using a Random Forest model on a rich set of features, we have obtained overall precision and recall scores of up to 0.92 and 0.72 respectively. These results show that the approach has validity, but we identify the type of focal seizure to be a critical factor for the classification performance.

AB - Epilepsy seizure detection with wearable devices is an emerging research field. As opposed to the gold standard, consisting of simultaneous video and EEG monitoring of patients, wearables have the advantage that they put a lower burden on epilepsy patients. We report on the first stages in a research effort that is dedicated to the development of a multimodal seizure detection system specifically for focal onset epileptic seizures. By in-depth analysis of data from three in-hospital patients with each having six to nine seizures recorded, we show that such seizures can manifest very differently and thus significantly impact classification. Using a Random Forest model on a rich set of features, we have obtained overall precision and recall scores of up to 0.92 and 0.72 respectively. These results show that the approach has validity, but we identify the type of focal seizure to be a critical factor for the classification performance.

KW - Epilepsy seizure detection

KW - Multimodal biosignals

KW - Wearables

UR - http://www.scopus.com/inward/record.url?scp=85078348577&partnerID=8YFLogxK

U2 - 10.1145/3361684.3361687

DO - 10.1145/3361684.3361687

M3 - Conference paper

AN - SCOPUS:85078348577

BT - iWOAR 2019 - 6th International Workshop on Sensor-Based Activity Recognition and Interaction, Proceedings

A2 - Ludtke, Stefan

A2 - Bader, Sebastian

A2 - Yordanova, Kristina

A2 - Kirste, Thomas

PB - Association for Computing Machinery

T2 - 6th International Workshop on Sensor-Based Activity Recognition and Interaction, iWOAR 2019

Y2 - 16 September 2019 through 17 September 2019

ER -

Böttcher S, Manyakov NV, Epitashvili N, Folarin A , Richardson MP, Dümpelmann M et al. Using multimodal biosignal data from wearables to detect focal motor seizures in individual epilepsy patients. In Ludtke S, Bader S, Yordanova K, Kirste T, editors, iWOAR 2019 - 6th International Workshop on Sensor-Based Activity Recognition and Interaction, Proceedings. Association for Computing Machinery. 2019. 3361687 doi: 10.1145/3361684.3361687

Using multimodal biosignal data from wearables to detect focal motor seizures in individual epilepsy patients

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this