Real-Time HR Estimation from wrist PPG using Binary LSTMs

Leandro Giacomini Rocha; Nick Van Helleputte; Muqing Liu; Dwaipayan Biswas; Bram Ernst Verhoef; Sergio Bampi; Chris H. Kim; Chris Van Hoof; Mario Konijnenburg; Marian Verhelst

doi:10.1109/BIOCAS.2019.8918726

Real-Time HR Estimation from wrist PPG using Binary LSTMs

Leandro Giacomini Rocha, Nick Van Helleputte, Muqing Liu, Dwaipayan Biswas, Bram Ernst Verhoef, Sergio Bampi, Chris H. Kim, Chris Van Hoof, Mario Konijnenburg, Marian Verhelst

Electrical and Computer Engineering

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

6 Scopus citations

Abstract

Wrist-worn photoplethysmography (PPG) sensors present a popular alternative to electrocardiogram recording for heart rate (HR) estimation. However, their accuracy is limited by motion artifacts inherent in ambulatory settings. In this paper, we propose a binarized neural network framework, b-CorNET, to efficiently estimate HR from single-channel wrist PPG signals during intense physical activity. The model comprises two binary convolution neural network layers followed by two binary long short-Term memory (b-LSTM) layers and a dense layer working on quantized PPG data. The proposed framework achieves an MAE of 3.75±3.05 bpm when evaluated on 12 IEEE SPC subjects. Furthermore, a novel, low-complexity architecture for the b-LSTM layers is proposed and efficiently mapped on a Xilinx Virtex5 FPGA, enabling HR computation.

Original language	English (US)
Title of host publication	BioCAS 2019 - Biomedical Circuits and Systems Conference, Proceedings
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781509006175
DOIs	https://doi.org/10.1109/BIOCAS.2019.8918726
State	Published - Oct 2019
Event	2019 IEEE Biomedical Circuits and Systems Conference, BioCAS 2019 - Nara, Japan Duration: Oct 17 2019 → Oct 19 2019

Publication series

Name	BioCAS 2019 - Biomedical Circuits and Systems Conference, Proceedings

Conference

Conference	2019 IEEE Biomedical Circuits and Systems Conference, BioCAS 2019
Country/Territory	Japan
City	Nara
Period	10/17/19 → 10/19/19

Bibliographical note

Funding Information:
V. CONCLUSION In this paper a b-CorNET framework with a proof-of-concept b-LSTM implementation is presented to estimate HR from wrist PPG data. The results in comparison to nonbinary CorNET show a slight degradation (< 0.1bpm), due to binarization and quantization. Future work will further increase the robustness of the model and integrate the CNN layers along with LSTM for a complete PPG HR estimation system. ACKNOWLEDGMENT The first author would like to thank IFRS, CAPES and CNPq Brazilian agencies for supporting this research. REFERENCES

Publisher Copyright:
© 2019 IEEE.

Keywords

Binary LSTM
CNN
FPGA
PPG

Access

10.1109/BIOCAS.2019.8918726

OpenUrl availability

Full text

Cite this

Rocha, L. G., Helleputte, N. V., Liu, M., Biswas, D., Verhoef, B. E., Bampi, S., Kim, C. H., Van Hoof, C., Konijnenburg, M., & Verhelst, M. (2019). Real-Time HR Estimation from wrist PPG using Binary LSTMs. In BioCAS 2019 - Biomedical Circuits and Systems Conference, Proceedings Article 8918726 (BioCAS 2019 - Biomedical Circuits and Systems Conference, Proceedings). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/BIOCAS.2019.8918726

Real-Time HR Estimation from wrist PPG using Binary LSTMs. / Rocha, Leandro Giacomini; Helleputte, Nick Van; Liu, Muqing et al.
BioCAS 2019 - Biomedical Circuits and Systems Conference, Proceedings. Institute of Electrical and Electronics Engineers Inc., 2019. 8918726 (BioCAS 2019 - Biomedical Circuits and Systems Conference, Proceedings).

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

Rocha, LG, Helleputte, NV, Liu, M, Biswas, D, Verhoef, BE, Bampi, S, Kim, CH, Van Hoof, C, Konijnenburg, M & Verhelst, M 2019, Real-Time HR Estimation from wrist PPG using Binary LSTMs. in BioCAS 2019 - Biomedical Circuits and Systems Conference, Proceedings., 8918726, BioCAS 2019 - Biomedical Circuits and Systems Conference, Proceedings, Institute of Electrical and Electronics Engineers Inc., 2019 IEEE Biomedical Circuits and Systems Conference, BioCAS 2019, Nara, Japan, 10/17/19. https://doi.org/10.1109/BIOCAS.2019.8918726

Rocha LG, Helleputte NV, Liu M, Biswas D, Verhoef BE, Bampi S et al. Real-Time HR Estimation from wrist PPG using Binary LSTMs. In BioCAS 2019 - Biomedical Circuits and Systems Conference, Proceedings. Institute of Electrical and Electronics Engineers Inc. 2019. 8918726. (BioCAS 2019 - Biomedical Circuits and Systems Conference, Proceedings). doi: 10.1109/BIOCAS.2019.8918726

@inproceedings{d86f169a5cf94b1c9b491b9f6ba022ac,

title = "Real-Time HR Estimation from wrist PPG using Binary LSTMs",

abstract = "Wrist-worn photoplethysmography (PPG) sensors present a popular alternative to electrocardiogram recording for heart rate (HR) estimation. However, their accuracy is limited by motion artifacts inherent in ambulatory settings. In this paper, we propose a binarized neural network framework, b-CorNET, to efficiently estimate HR from single-channel wrist PPG signals during intense physical activity. The model comprises two binary convolution neural network layers followed by two binary long short-Term memory (b-LSTM) layers and a dense layer working on quantized PPG data. The proposed framework achieves an MAE of 3.75±3.05 bpm when evaluated on 12 IEEE SPC subjects. Furthermore, a novel, low-complexity architecture for the b-LSTM layers is proposed and efficiently mapped on a Xilinx Virtex5 FPGA, enabling HR computation.",

keywords = "Binary LSTM, CNN, FPGA, PPG",

author = "Rocha, {Leandro Giacomini} and Helleputte, {Nick Van} and Muqing Liu and Dwaipayan Biswas and Verhoef, {Bram Ernst} and Sergio Bampi and Kim, {Chris H.} and {Van Hoof}, Chris and Mario Konijnenburg and Marian Verhelst",

note = "Funding Information: V. CONCLUSION In this paper a b-CorNET framework with a proof-of-concept b-LSTM implementation is presented to estimate HR from wrist PPG data. The results in comparison to nonbinary CorNET show a slight degradation (< 0.1bpm), due to binarization and quantization. Future work will further increase the robustness of the model and integrate the CNN layers along with LSTM for a complete PPG HR estimation system. ACKNOWLEDGMENT The first author would like to thank IFRS, CAPES and CNPq Brazilian agencies for supporting this research. REFERENCES Publisher Copyright: {\textcopyright} 2019 IEEE.; 2019 IEEE Biomedical Circuits and Systems Conference, BioCAS 2019 ; Conference date: 17-10-2019 Through 19-10-2019",

year = "2019",

month = oct,

doi = "10.1109/BIOCAS.2019.8918726",

language = "English (US)",

series = "BioCAS 2019 - Biomedical Circuits and Systems Conference, Proceedings",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "BioCAS 2019 - Biomedical Circuits and Systems Conference, Proceedings",

}

TY - GEN

T1 - Real-Time HR Estimation from wrist PPG using Binary LSTMs

AU - Rocha, Leandro Giacomini

AU - Helleputte, Nick Van

AU - Liu, Muqing

AU - Biswas, Dwaipayan

AU - Verhoef, Bram Ernst

AU - Bampi, Sergio

AU - Kim, Chris H.

AU - Van Hoof, Chris

AU - Konijnenburg, Mario

AU - Verhelst, Marian

N1 - Funding Information: V. CONCLUSION In this paper a b-CorNET framework with a proof-of-concept b-LSTM implementation is presented to estimate HR from wrist PPG data. The results in comparison to nonbinary CorNET show a slight degradation (< 0.1bpm), due to binarization and quantization. Future work will further increase the robustness of the model and integrate the CNN layers along with LSTM for a complete PPG HR estimation system. ACKNOWLEDGMENT The first author would like to thank IFRS, CAPES and CNPq Brazilian agencies for supporting this research. REFERENCES Publisher Copyright: © 2019 IEEE.

PY - 2019/10

Y1 - 2019/10

N2 - Wrist-worn photoplethysmography (PPG) sensors present a popular alternative to electrocardiogram recording for heart rate (HR) estimation. However, their accuracy is limited by motion artifacts inherent in ambulatory settings. In this paper, we propose a binarized neural network framework, b-CorNET, to efficiently estimate HR from single-channel wrist PPG signals during intense physical activity. The model comprises two binary convolution neural network layers followed by two binary long short-Term memory (b-LSTM) layers and a dense layer working on quantized PPG data. The proposed framework achieves an MAE of 3.75±3.05 bpm when evaluated on 12 IEEE SPC subjects. Furthermore, a novel, low-complexity architecture for the b-LSTM layers is proposed and efficiently mapped on a Xilinx Virtex5 FPGA, enabling HR computation.

AB - Wrist-worn photoplethysmography (PPG) sensors present a popular alternative to electrocardiogram recording for heart rate (HR) estimation. However, their accuracy is limited by motion artifacts inherent in ambulatory settings. In this paper, we propose a binarized neural network framework, b-CorNET, to efficiently estimate HR from single-channel wrist PPG signals during intense physical activity. The model comprises two binary convolution neural network layers followed by two binary long short-Term memory (b-LSTM) layers and a dense layer working on quantized PPG data. The proposed framework achieves an MAE of 3.75±3.05 bpm when evaluated on 12 IEEE SPC subjects. Furthermore, a novel, low-complexity architecture for the b-LSTM layers is proposed and efficiently mapped on a Xilinx Virtex5 FPGA, enabling HR computation.

KW - Binary LSTM

KW - CNN

KW - FPGA

KW - PPG

UR - http://www.scopus.com/inward/record.url?scp=85077044436&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85077044436&partnerID=8YFLogxK

U2 - 10.1109/BIOCAS.2019.8918726

DO - 10.1109/BIOCAS.2019.8918726

M3 - Conference contribution

AN - SCOPUS:85077044436

T3 - BioCAS 2019 - Biomedical Circuits and Systems Conference, Proceedings

BT - BioCAS 2019 - Biomedical Circuits and Systems Conference, Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2019 IEEE Biomedical Circuits and Systems Conference, BioCAS 2019

Y2 - 17 October 2019 through 19 October 2019

ER -

Real-Time HR Estimation from wrist PPG using Binary LSTMs

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access

OpenUrl availability

Other files and links

Fingerprint

Cite this