Learning to predict coronary perfusion pressure during cardiopulmonary resuscitation

Manan Gandhi; Yunpeng Pan; Evangelos Theodorou; Pierre Sebastian; Matt Olson; Demetri Yannopoulos

doi:10.1115/DSCC2018-8968

Learning to predict coronary perfusion pressure during cardiopulmonary resuscitation

Manan Gandhi, Yunpeng Pan, Evangelos Theodorou, Pierre Sebastian, Matt Olson, Demetri Yannopoulos

Medicine - Cardiology Division

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

Abstract

The goal of this work is to advance the capability of automated, mechanical cardiopulmonary resuscitation (CPR) by predicting Coronary Perfusion Pressure (CPP) within 5 mmHg at a given moment in time. We aim to utilize methods from machine learning in order to model the CPP of a porcine patient subjected to automated chest compressions. During preprocessing of the data, we show how data sampling rate, delays and moving average filtering can improve predictions. We demonstrate state of the art modeling performance utilizing a variety of algorithms, and analyze the performance of each algorithm for single-step and long-term predictions. The results indicate that a delayed linear system achieves this target CPP within 0.25 mmHg. For longer time horizons, a more complex model is required. We demonstrate that the Long-short-term-memory (LSTM) network has the best single run performance, while the Sparse Spectrum Gaussian Process (SSGP) has the best average performance.

Original language	English (US)
Title of host publication	Advances in Control Design Methods; Advances in Nonlinear Control; Advances in Robotics; Assistive and Rehabilitation Robotics; Automotive Dynamics and Emerging Powertrain Technologies; Automotive Systems; Bio Engineering Applications; Bio-Mechatronics and Physical Human Robot Interaction; Biomedical and Neural Systems; Biomedical and Neural Systems Modeling, Diagnostics, and Healthcare
Publisher	American Society of Mechanical Engineers (ASME)
ISBN (Electronic)	9780791851890
DOIs	https://doi.org/10.1115/DSCC2018-8968
State	Published - 2018
Event	ASME 2018 Dynamic Systems and Control Conference, DSCC 2018 - Atlanta, United States Duration: Sep 30 2018 → Oct 3 2018

Publication series

Name	ASME 2018 Dynamic Systems and Control Conference, DSCC 2018
Volume	1

Other

Other	ASME 2018 Dynamic Systems and Control Conference, DSCC 2018
Country/Territory	United States
City	Atlanta
Period	9/30/18 → 10/3/18

Bibliographical note

Funding Information:
This work was supported by the Department of Defense (DoD) through the National Defense Science & Engineering Graduate Fellowship (NDSEG) Program.

Publisher Copyright:
Copyright © 2018 ASME.

Access

10.1115/DSCC2018-8968

OpenUrl availability

Full text

Cite this

Gandhi, M., Pan, Y., Theodorou, E., Sebastian, P., Olson, M., & Yannopoulos, D. (2018). Learning to predict coronary perfusion pressure during cardiopulmonary resuscitation. In Advances in Control Design Methods; Advances in Nonlinear Control; Advances in Robotics; Assistive and Rehabilitation Robotics; Automotive Dynamics and Emerging Powertrain Technologies; Automotive Systems; Bio Engineering Applications; Bio-Mechatronics and Physical Human Robot Interaction; Biomedical and Neural Systems; Biomedical and Neural Systems Modeling, Diagnostics, and Healthcare (ASME 2018 Dynamic Systems and Control Conference, DSCC 2018; Vol. 1). American Society of Mechanical Engineers (ASME). https://doi.org/10.1115/DSCC2018-8968

Learning to predict coronary perfusion pressure during cardiopulmonary resuscitation. / Gandhi, Manan; Pan, Yunpeng; Theodorou, Evangelos et al.
Advances in Control Design Methods; Advances in Nonlinear Control; Advances in Robotics; Assistive and Rehabilitation Robotics; Automotive Dynamics and Emerging Powertrain Technologies; Automotive Systems; Bio Engineering Applications; Bio-Mechatronics and Physical Human Robot Interaction; Biomedical and Neural Systems; Biomedical and Neural Systems Modeling, Diagnostics, and Healthcare. American Society of Mechanical Engineers (ASME), 2018. (ASME 2018 Dynamic Systems and Control Conference, DSCC 2018; Vol. 1).

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

Gandhi, M, Pan, Y, Theodorou, E, Sebastian, P, Olson, M & Yannopoulos, D 2018, Learning to predict coronary perfusion pressure during cardiopulmonary resuscitation. in Advances in Control Design Methods; Advances in Nonlinear Control; Advances in Robotics; Assistive and Rehabilitation Robotics; Automotive Dynamics and Emerging Powertrain Technologies; Automotive Systems; Bio Engineering Applications; Bio-Mechatronics and Physical Human Robot Interaction; Biomedical and Neural Systems; Biomedical and Neural Systems Modeling, Diagnostics, and Healthcare. ASME 2018 Dynamic Systems and Control Conference, DSCC 2018, vol. 1, American Society of Mechanical Engineers (ASME), ASME 2018 Dynamic Systems and Control Conference, DSCC 2018, Atlanta, United States, 9/30/18. https://doi.org/10.1115/DSCC2018-8968

Gandhi M, Pan Y, Theodorou E, Sebastian P, Olson M, Yannopoulos D. Learning to predict coronary perfusion pressure during cardiopulmonary resuscitation. In Advances in Control Design Methods; Advances in Nonlinear Control; Advances in Robotics; Assistive and Rehabilitation Robotics; Automotive Dynamics and Emerging Powertrain Technologies; Automotive Systems; Bio Engineering Applications; Bio-Mechatronics and Physical Human Robot Interaction; Biomedical and Neural Systems; Biomedical and Neural Systems Modeling, Diagnostics, and Healthcare. American Society of Mechanical Engineers (ASME). 2018. (ASME 2018 Dynamic Systems and Control Conference, DSCC 2018). doi: 10.1115/DSCC2018-8968

Gandhi, Manan ; Pan, Yunpeng ; Theodorou, Evangelos et al. / Learning to predict coronary perfusion pressure during cardiopulmonary resuscitation. Advances in Control Design Methods; Advances in Nonlinear Control; Advances in Robotics; Assistive and Rehabilitation Robotics; Automotive Dynamics and Emerging Powertrain Technologies; Automotive Systems; Bio Engineering Applications; Bio-Mechatronics and Physical Human Robot Interaction; Biomedical and Neural Systems; Biomedical and Neural Systems Modeling, Diagnostics, and Healthcare. American Society of Mechanical Engineers (ASME), 2018. (ASME 2018 Dynamic Systems and Control Conference, DSCC 2018).

@inproceedings{e6d78afd9cfd41179c0789463d9faf91,

title = "Learning to predict coronary perfusion pressure during cardiopulmonary resuscitation",

abstract = "The goal of this work is to advance the capability of automated, mechanical cardiopulmonary resuscitation (CPR) by predicting Coronary Perfusion Pressure (CPP) within 5 mmHg at a given moment in time. We aim to utilize methods from machine learning in order to model the CPP of a porcine patient subjected to automated chest compressions. During preprocessing of the data, we show how data sampling rate, delays and moving average filtering can improve predictions. We demonstrate state of the art modeling performance utilizing a variety of algorithms, and analyze the performance of each algorithm for single-step and long-term predictions. The results indicate that a delayed linear system achieves this target CPP within 0.25 mmHg. For longer time horizons, a more complex model is required. We demonstrate that the Long-short-term-memory (LSTM) network has the best single run performance, while the Sparse Spectrum Gaussian Process (SSGP) has the best average performance.",

author = "Manan Gandhi and Yunpeng Pan and Evangelos Theodorou and Pierre Sebastian and Matt Olson and Demetri Yannopoulos",

note = "Funding Information: This work was supported by the Department of Defense (DoD) through the National Defense Science & Engineering Graduate Fellowship (NDSEG) Program. Publisher Copyright: Copyright {\textcopyright} 2018 ASME.; ASME 2018 Dynamic Systems and Control Conference, DSCC 2018 ; Conference date: 30-09-2018 Through 03-10-2018",

year = "2018",

doi = "10.1115/DSCC2018-8968",

language = "English (US)",

series = "ASME 2018 Dynamic Systems and Control Conference, DSCC 2018",

publisher = "American Society of Mechanical Engineers (ASME)",

booktitle = "Advances in Control Design Methods; Advances in Nonlinear Control; Advances in Robotics; Assistive and Rehabilitation Robotics; Automotive Dynamics and Emerging Powertrain Technologies; Automotive Systems; Bio Engineering Applications; Bio-Mechatronics and Physical Human Robot Interaction; Biomedical and Neural Systems; Biomedical and Neural Systems Modeling, Diagnostics, and Healthcare",

}

TY - GEN

T1 - Learning to predict coronary perfusion pressure during cardiopulmonary resuscitation

AU - Gandhi, Manan

AU - Pan, Yunpeng

AU - Theodorou, Evangelos

AU - Sebastian, Pierre

AU - Olson, Matt

AU - Yannopoulos, Demetri

N1 - Funding Information: This work was supported by the Department of Defense (DoD) through the National Defense Science & Engineering Graduate Fellowship (NDSEG) Program. Publisher Copyright: Copyright © 2018 ASME.

PY - 2018

Y1 - 2018

N2 - The goal of this work is to advance the capability of automated, mechanical cardiopulmonary resuscitation (CPR) by predicting Coronary Perfusion Pressure (CPP) within 5 mmHg at a given moment in time. We aim to utilize methods from machine learning in order to model the CPP of a porcine patient subjected to automated chest compressions. During preprocessing of the data, we show how data sampling rate, delays and moving average filtering can improve predictions. We demonstrate state of the art modeling performance utilizing a variety of algorithms, and analyze the performance of each algorithm for single-step and long-term predictions. The results indicate that a delayed linear system achieves this target CPP within 0.25 mmHg. For longer time horizons, a more complex model is required. We demonstrate that the Long-short-term-memory (LSTM) network has the best single run performance, while the Sparse Spectrum Gaussian Process (SSGP) has the best average performance.

AB - The goal of this work is to advance the capability of automated, mechanical cardiopulmonary resuscitation (CPR) by predicting Coronary Perfusion Pressure (CPP) within 5 mmHg at a given moment in time. We aim to utilize methods from machine learning in order to model the CPP of a porcine patient subjected to automated chest compressions. During preprocessing of the data, we show how data sampling rate, delays and moving average filtering can improve predictions. We demonstrate state of the art modeling performance utilizing a variety of algorithms, and analyze the performance of each algorithm for single-step and long-term predictions. The results indicate that a delayed linear system achieves this target CPP within 0.25 mmHg. For longer time horizons, a more complex model is required. We demonstrate that the Long-short-term-memory (LSTM) network has the best single run performance, while the Sparse Spectrum Gaussian Process (SSGP) has the best average performance.

UR - http://www.scopus.com/inward/record.url?scp=85057323305&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85057323305&partnerID=8YFLogxK

U2 - 10.1115/DSCC2018-8968

DO - 10.1115/DSCC2018-8968

M3 - Conference contribution

AN - SCOPUS:85057323305

T3 - ASME 2018 Dynamic Systems and Control Conference, DSCC 2018

BT - Advances in Control Design Methods; Advances in Nonlinear Control; Advances in Robotics; Assistive and Rehabilitation Robotics; Automotive Dynamics and Emerging Powertrain Technologies; Automotive Systems; Bio Engineering Applications; Bio-Mechatronics and Physical Human Robot Interaction; Biomedical and Neural Systems; Biomedical and Neural Systems Modeling, Diagnostics, and Healthcare

PB - American Society of Mechanical Engineers (ASME)

T2 - ASME 2018 Dynamic Systems and Control Conference, DSCC 2018

Y2 - 30 September 2018 through 3 October 2018

ER -

Learning to predict coronary perfusion pressure during cardiopulmonary resuscitation

Abstract

Publication series

Other

Bibliographical note

Access

OpenUrl availability

Other files and links

Fingerprint

Cite this