CorNET: Deep Learning Framework for PPG-Based Heart Rate Estimation and Biometric Identification in Ambulant Environment

Dwaipayan Biswas, Luke R Everson, Muqing Liu, Madhuri Panwar, Bram Ernst Verhoef, Shrishail Patki, Chris H. Kim, Amit Acharyya, Chris Van Hoof, Mario Konijnenburg, Nick Van Helleputte

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

Advancements in wireless sensor network technologies have enabled the proliferation of miniaturized body-worn sensors, capable of long-Term pervasive biomedical signal monitoring. Remote cardiovascular monitoring has been one of the beneficiaries of this development, resulting in non-invasive, photoplethysmography (PPG) sensors being used in ambulatory settings. Wrist-worn PPG, although a popular alternative to electrocardiogram, suffers from motion artifacts inherent in daily life. Hence, in this paper, we present a novel deep learning framework (CorNET) to efficiently estimate heart rate (HR) information and perform biometric identification (BId) using only a wrist-worn, single-channel PPG signal collected in ambulant environment. We have formulated a completely personalized data-driven approach, using a four-layer deep neural network. Two convolution neural network layers are used in conjunction with two long short-Term memory layers, followed by a dense output layer for modeling the temporal sequence inherent within the pulsatile signal representative of cardiac activity. The final dense layer is customized with respect to the application, functioning as: regression layer-having a single neuron to predict HR; classification layer-two neurons that identify a subject among a group. The proposed network was evaluated on the TROIKA dataset having 22 PPG records collected during various physical activities. We achieve a mean absolute error of 1.47 ± 3.37 beats per minute for HR estimation and an average accuracy of 96% for BId on 20 subjects. CorNET was further evaluated successfully in an ambulant use-case scenario with custom sensors for two subjects.

Original languageEnglish (US)
Article number8607019
Pages (from-to)282-291
Number of pages10
JournalIEEE transactions on biomedical circuits and systems
Volume13
Issue number2
DOIs
StatePublished - Apr 1 2019

Fingerprint

Photoplethysmography
Biometrics
Neurons
Sensors
Monitoring
Network layers
Electrocardiography
Convolution
Wireless sensor networks
Neural networks
Deep learning

Keywords

  • Average heart rate
  • PPG
  • biometric
  • convolutional neural network
  • deep learning
  • long short-Term memory

PubMed: MeSH publication types

  • Journal Article
  • Research Support, U.S. Gov't, Non-P.H.S.

Cite this

CorNET : Deep Learning Framework for PPG-Based Heart Rate Estimation and Biometric Identification in Ambulant Environment. / Biswas, Dwaipayan; Everson, Luke R; Liu, Muqing; Panwar, Madhuri; Verhoef, Bram Ernst; Patki, Shrishail; Kim, Chris H.; Acharyya, Amit; Van Hoof, Chris; Konijnenburg, Mario; Van Helleputte, Nick.

In: IEEE transactions on biomedical circuits and systems, Vol. 13, No. 2, 8607019, 01.04.2019, p. 282-291.

Research output: Contribution to journalArticle

Biswas, D, Everson, LR, Liu, M, Panwar, M, Verhoef, BE, Patki, S, Kim, CH, Acharyya, A, Van Hoof, C, Konijnenburg, M & Van Helleputte, N 2019, 'CorNET: Deep Learning Framework for PPG-Based Heart Rate Estimation and Biometric Identification in Ambulant Environment', IEEE transactions on biomedical circuits and systems, vol. 13, no. 2, 8607019, pp. 282-291. https://doi.org/10.1109/TBCAS.2019.2892297
Biswas, Dwaipayan ; Everson, Luke R ; Liu, Muqing ; Panwar, Madhuri ; Verhoef, Bram Ernst ; Patki, Shrishail ; Kim, Chris H. ; Acharyya, Amit ; Van Hoof, Chris ; Konijnenburg, Mario ; Van Helleputte, Nick. / CorNET : Deep Learning Framework for PPG-Based Heart Rate Estimation and Biometric Identification in Ambulant Environment. In: IEEE transactions on biomedical circuits and systems. 2019 ; Vol. 13, No. 2. pp. 282-291.
@article{89b152628f814f1289e8819d1989cd63,
title = "CorNET: Deep Learning Framework for PPG-Based Heart Rate Estimation and Biometric Identification in Ambulant Environment",
abstract = "Advancements in wireless sensor network technologies have enabled the proliferation of miniaturized body-worn sensors, capable of long-Term pervasive biomedical signal monitoring. Remote cardiovascular monitoring has been one of the beneficiaries of this development, resulting in non-invasive, photoplethysmography (PPG) sensors being used in ambulatory settings. Wrist-worn PPG, although a popular alternative to electrocardiogram, suffers from motion artifacts inherent in daily life. Hence, in this paper, we present a novel deep learning framework (CorNET) to efficiently estimate heart rate (HR) information and perform biometric identification (BId) using only a wrist-worn, single-channel PPG signal collected in ambulant environment. We have formulated a completely personalized data-driven approach, using a four-layer deep neural network. Two convolution neural network layers are used in conjunction with two long short-Term memory layers, followed by a dense output layer for modeling the temporal sequence inherent within the pulsatile signal representative of cardiac activity. The final dense layer is customized with respect to the application, functioning as: regression layer-having a single neuron to predict HR; classification layer-two neurons that identify a subject among a group. The proposed network was evaluated on the TROIKA dataset having 22 PPG records collected during various physical activities. We achieve a mean absolute error of 1.47 ± 3.37 beats per minute for HR estimation and an average accuracy of 96{\%} for BId on 20 subjects. CorNET was further evaluated successfully in an ambulant use-case scenario with custom sensors for two subjects.",
keywords = "Average heart rate, PPG, biometric, convolutional neural network, deep learning, long short-Term memory",
author = "Dwaipayan Biswas and Everson, {Luke R} and Muqing Liu and Madhuri Panwar and Verhoef, {Bram Ernst} and Shrishail Patki and Kim, {Chris H.} and Amit Acharyya and {Van Hoof}, Chris and Mario Konijnenburg and {Van Helleputte}, Nick",
year = "2019",
month = "4",
day = "1",
doi = "10.1109/TBCAS.2019.2892297",
language = "English (US)",
volume = "13",
pages = "282--291",
journal = "IEEE Transactions on Biomedical Circuits and Systems",
issn = "1932-4545",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "2",

}

TY - JOUR

T1 - CorNET

T2 - Deep Learning Framework for PPG-Based Heart Rate Estimation and Biometric Identification in Ambulant Environment

AU - Biswas, Dwaipayan

AU - Everson, Luke R

AU - Liu, Muqing

AU - Panwar, Madhuri

AU - Verhoef, Bram Ernst

AU - Patki, Shrishail

AU - Kim, Chris H.

AU - Acharyya, Amit

AU - Van Hoof, Chris

AU - Konijnenburg, Mario

AU - Van Helleputte, Nick

PY - 2019/4/1

Y1 - 2019/4/1

N2 - Advancements in wireless sensor network technologies have enabled the proliferation of miniaturized body-worn sensors, capable of long-Term pervasive biomedical signal monitoring. Remote cardiovascular monitoring has been one of the beneficiaries of this development, resulting in non-invasive, photoplethysmography (PPG) sensors being used in ambulatory settings. Wrist-worn PPG, although a popular alternative to electrocardiogram, suffers from motion artifacts inherent in daily life. Hence, in this paper, we present a novel deep learning framework (CorNET) to efficiently estimate heart rate (HR) information and perform biometric identification (BId) using only a wrist-worn, single-channel PPG signal collected in ambulant environment. We have formulated a completely personalized data-driven approach, using a four-layer deep neural network. Two convolution neural network layers are used in conjunction with two long short-Term memory layers, followed by a dense output layer for modeling the temporal sequence inherent within the pulsatile signal representative of cardiac activity. The final dense layer is customized with respect to the application, functioning as: regression layer-having a single neuron to predict HR; classification layer-two neurons that identify a subject among a group. The proposed network was evaluated on the TROIKA dataset having 22 PPG records collected during various physical activities. We achieve a mean absolute error of 1.47 ± 3.37 beats per minute for HR estimation and an average accuracy of 96% for BId on 20 subjects. CorNET was further evaluated successfully in an ambulant use-case scenario with custom sensors for two subjects.

AB - Advancements in wireless sensor network technologies have enabled the proliferation of miniaturized body-worn sensors, capable of long-Term pervasive biomedical signal monitoring. Remote cardiovascular monitoring has been one of the beneficiaries of this development, resulting in non-invasive, photoplethysmography (PPG) sensors being used in ambulatory settings. Wrist-worn PPG, although a popular alternative to electrocardiogram, suffers from motion artifacts inherent in daily life. Hence, in this paper, we present a novel deep learning framework (CorNET) to efficiently estimate heart rate (HR) information and perform biometric identification (BId) using only a wrist-worn, single-channel PPG signal collected in ambulant environment. We have formulated a completely personalized data-driven approach, using a four-layer deep neural network. Two convolution neural network layers are used in conjunction with two long short-Term memory layers, followed by a dense output layer for modeling the temporal sequence inherent within the pulsatile signal representative of cardiac activity. The final dense layer is customized with respect to the application, functioning as: regression layer-having a single neuron to predict HR; classification layer-two neurons that identify a subject among a group. The proposed network was evaluated on the TROIKA dataset having 22 PPG records collected during various physical activities. We achieve a mean absolute error of 1.47 ± 3.37 beats per minute for HR estimation and an average accuracy of 96% for BId on 20 subjects. CorNET was further evaluated successfully in an ambulant use-case scenario with custom sensors for two subjects.

KW - Average heart rate

KW - PPG

KW - biometric

KW - convolutional neural network

KW - deep learning

KW - long short-Term memory

UR - http://www.scopus.com/inward/record.url?scp=85063429157&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85063429157&partnerID=8YFLogxK

U2 - 10.1109/TBCAS.2019.2892297

DO - 10.1109/TBCAS.2019.2892297

M3 - Article

C2 - 30629514

AN - SCOPUS:85063429157

VL - 13

SP - 282

EP - 291

JO - IEEE Transactions on Biomedical Circuits and Systems

JF - IEEE Transactions on Biomedical Circuits and Systems

SN - 1932-4545

IS - 2

M1 - 8607019

ER -