Estimation of cardiac output based on PRAM algorithm and ARX model from noninvasive radial pressure wave

Lu Wang; Lisheng Xu; Yu Sun; Kai Xu; Libo Zhang; Steve Greenwald; Dingchang Zheng

doi:10.1016/j.bspc.2025.108984

Estimation of cardiac output based on PRAM algorithm and ARX model from noninvasive radial pressure wave

Lu Wang
, Lisheng Xu
, Yu Sun
, Kai Xu
, Libo Zhang
, Steve Greenwald
, Dingchang Zheng

Northeastern University
Ministry of Education China
Key Laboratory of Cardiovascular Imaging and Research of Liaoning Province
General Hospital of Northern Theater Command
University of London

Research output: Contribution to journal › Article › peer-review

38 Downloads (Pure)

Abstract

There is an increasing need for the prevention, diagnosis and treatment of cardiovascular disease (CVD). Reduced cardiac output (CO) is frequently associated with CVD. Thus, accurate measurement of CO aids its diagnosis and helps to guide treatment. A convenient, real-time, noninvasive way to estimate CO is the Pressure Recording Analytic Method (PRAM). It is based on the non-invasive detection of the pressure wave in a peripheral artery and does not require invasive measurements to calibrate the model parameters. Unfortunately, its accuracy is limited due to its dependence on the patient's height, weight, heart rate, and the mechanical properties of each individual's arteries. Furthermore, compared to the peripheral arterial pressure wave, the aortic pressure wave provides a more accurate and efficient means of estimating CO. Therefore, to improve the accuracy of the original PRAM method, this study incorporates height, weight, and heart rate measurements, as well as an Auto-Regressive with eXogenous input (ARX) model, enabling adaptive estimation of the aortic pressure waveform from radial artery pressure wave measurement. The CO estimations of the original peripheral PRAM (CO_PRAMper), the improved peripheral PRAM (CO_IPRAMper) and the improved central PRAM (CO_IPRAMcen) were compared to MRI results (CO_MRI), as the ground truth. The correlation coefficients (R²) between the CO estimates using the 3 algorithms and CO_MRI were 0.271, 0.548 and 0.757, respectively. These R² values were statistically significant and showed that CO_IPRAMcen performed best. The mean difference between the CO estimates using the 3 algorithms and CO_MRI were −0.15 ± 0.44, −0.07 ± 0.24 and −0.04 ± 0.17 L/min, respectively.

Original language	English
Article number	108984
Number of pages	10
Journal	Biomedical Signal Processing and Control
Volume	113
Issue number	Part A
Early online date	30 Oct 2025
DOIs	https://doi.org/10.1016/j.bspc.2025.108984
Publication status	Published - Mar 2026

Bibliographical note

Publisher Copyright:
© 2025 Elsevier Ltd
Copyright © and Moral Rights are retained by the author(s) and/ or other copyright owners. A copy can be downloaded for personal non-commercial research or study, without prior permission or charge. This item cannot be reproduced or quoted extensively from without first obtaining permission in writing from the copyright holder(s). The content must not be changed in any way or sold commercially in any format or medium without the formal permission of the copyright holders.

This document is the author’s post-print version, incorporating any revisions agreed during the peer-review process. Some differences between the published version and this version may remain and you are advised to consult the published version if you wish to cite from it

Funding

This work was supported by the National Natural Science Foundation of China (Nos. 62273082, and 61773110), the Natural Science Foundation of Liaoning Province (No. 2021-YGJC-14), the Liaoning Province Science and Technology Plan Project (No. 2023JH2/101300125), and the Fundamental Research Funds for the Central Universities (N25ZLL045).

Funders	Funder number
National Natural Science Foundation of China	62273082, 61773110
Natural Science Foundation of Liaoning Province	2021-YGJC-14
Fundamental Research Funds for the Central Universities	N25ZLL045
Liaoning Provincial Science and Technology Program	2023JH2/101300125

UN SDGs

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

SDG 3 Good Health and Well-being

Keywords

Aortic pressure wave
Cardiac output
MRI
Peripheral pressure wave
PRAM

ASJC Scopus subject areas

Signal Processing
Biomedical Engineering
Health Informatics

Access to Document

10.1016/j.bspc.2025.108984

Zheng2025AAMAccepted author manuscript, 693 KBLicence: CC BY-NC-ND

Cite this

@article{9b2d8c3b93704eef8a590991fe72a0d2,

title = "Estimation of cardiac output based on PRAM algorithm and ARX model from noninvasive radial pressure wave",

abstract = "There is an increasing need for the prevention, diagnosis and treatment of cardiovascular disease (CVD). Reduced cardiac output (CO) is frequently associated with CVD. Thus, accurate measurement of CO aids its diagnosis and helps to guide treatment. A convenient, real-time, noninvasive way to estimate CO is the Pressure Recording Analytic Method (PRAM). It is based on the non-invasive detection of the pressure wave in a peripheral artery and does not require invasive measurements to calibrate the model parameters. Unfortunately, its accuracy is limited due to its dependence on the patient's height, weight, heart rate, and the mechanical properties of each individual's arteries. Furthermore, compared to the peripheral arterial pressure wave, the aortic pressure wave provides a more accurate and efficient means of estimating CO. Therefore, to improve the accuracy of the original PRAM method, this study incorporates height, weight, and heart rate measurements, as well as an Auto-Regressive with eXogenous input (ARX) model, enabling adaptive estimation of the aortic pressure waveform from radial artery pressure wave measurement. The CO estimations of the original peripheral PRAM (COPRAMper), the improved peripheral PRAM (COIPRAMper) and the improved central PRAM (COIPRAMcen) were compared to MRI results (COMRI), as the ground truth. The correlation coefficients (R2) between the CO estimates using the 3 algorithms and COMRI were 0.271, 0.548 and 0.757, respectively. These R2 values were statistically significant and showed that COIPRAMcen performed best. The mean difference between the CO estimates using the 3 algorithms and COMRI were −0.15 ± 0.44, −0.07 ± 0.24 and −0.04 ± 0.17 L/min, respectively.",

keywords = "Aortic pressure wave, Cardiac output, MRI, Peripheral pressure wave, PRAM",

author = "Lu Wang and Lisheng Xu and Yu Sun and Kai Xu and Libo Zhang and Steve Greenwald and Dingchang Zheng",

note = "Publisher Copyright: {\textcopyright} 2025 Elsevier Ltd Copyright {\textcopyright} and Moral Rights are retained by the author(s) and/ or other copyright owners. A copy can be downloaded for personal non-commercial research or study, without prior permission or charge. This item cannot be reproduced or quoted extensively from without first obtaining permission in writing from the copyright holder(s). The content must not be changed in any way or sold commercially in any format or medium without the formal permission of the copyright holders. This document is the author{\textquoteright}s post-print version, incorporating any revisions agreed during the peer-review process. Some differences between the published version and this version may remain and you are advised to consult the published version if you wish to cite from it ",

year = "2026",

month = mar,

doi = "10.1016/j.bspc.2025.108984",

language = "English",

volume = "113",

journal = "Biomedical Signal Processing and Control",

issn = "1746-8094",

publisher = "Elsevier",

number = "Part A",

}

TY - JOUR

T1 - Estimation of cardiac output based on PRAM algorithm and ARX model from noninvasive radial pressure wave

AU - Wang, Lu

AU - Xu, Lisheng

AU - Sun, Yu

AU - Xu, Kai

AU - Zhang, Libo

AU - Greenwald, Steve

AU - Zheng, Dingchang

N1 - Publisher Copyright: © 2025 Elsevier Ltd Copyright © and Moral Rights are retained by the author(s) and/ or other copyright owners. A copy can be downloaded for personal non-commercial research or study, without prior permission or charge. This item cannot be reproduced or quoted extensively from without first obtaining permission in writing from the copyright holder(s). The content must not be changed in any way or sold commercially in any format or medium without the formal permission of the copyright holders. This document is the author’s post-print version, incorporating any revisions agreed during the peer-review process. Some differences between the published version and this version may remain and you are advised to consult the published version if you wish to cite from it

PY - 2026/3

Y1 - 2026/3

N2 - There is an increasing need for the prevention, diagnosis and treatment of cardiovascular disease (CVD). Reduced cardiac output (CO) is frequently associated with CVD. Thus, accurate measurement of CO aids its diagnosis and helps to guide treatment. A convenient, real-time, noninvasive way to estimate CO is the Pressure Recording Analytic Method (PRAM). It is based on the non-invasive detection of the pressure wave in a peripheral artery and does not require invasive measurements to calibrate the model parameters. Unfortunately, its accuracy is limited due to its dependence on the patient's height, weight, heart rate, and the mechanical properties of each individual's arteries. Furthermore, compared to the peripheral arterial pressure wave, the aortic pressure wave provides a more accurate and efficient means of estimating CO. Therefore, to improve the accuracy of the original PRAM method, this study incorporates height, weight, and heart rate measurements, as well as an Auto-Regressive with eXogenous input (ARX) model, enabling adaptive estimation of the aortic pressure waveform from radial artery pressure wave measurement. The CO estimations of the original peripheral PRAM (COPRAMper), the improved peripheral PRAM (COIPRAMper) and the improved central PRAM (COIPRAMcen) were compared to MRI results (COMRI), as the ground truth. The correlation coefficients (R2) between the CO estimates using the 3 algorithms and COMRI were 0.271, 0.548 and 0.757, respectively. These R2 values were statistically significant and showed that COIPRAMcen performed best. The mean difference between the CO estimates using the 3 algorithms and COMRI were −0.15 ± 0.44, −0.07 ± 0.24 and −0.04 ± 0.17 L/min, respectively.

AB - There is an increasing need for the prevention, diagnosis and treatment of cardiovascular disease (CVD). Reduced cardiac output (CO) is frequently associated with CVD. Thus, accurate measurement of CO aids its diagnosis and helps to guide treatment. A convenient, real-time, noninvasive way to estimate CO is the Pressure Recording Analytic Method (PRAM). It is based on the non-invasive detection of the pressure wave in a peripheral artery and does not require invasive measurements to calibrate the model parameters. Unfortunately, its accuracy is limited due to its dependence on the patient's height, weight, heart rate, and the mechanical properties of each individual's arteries. Furthermore, compared to the peripheral arterial pressure wave, the aortic pressure wave provides a more accurate and efficient means of estimating CO. Therefore, to improve the accuracy of the original PRAM method, this study incorporates height, weight, and heart rate measurements, as well as an Auto-Regressive with eXogenous input (ARX) model, enabling adaptive estimation of the aortic pressure waveform from radial artery pressure wave measurement. The CO estimations of the original peripheral PRAM (COPRAMper), the improved peripheral PRAM (COIPRAMper) and the improved central PRAM (COIPRAMcen) were compared to MRI results (COMRI), as the ground truth. The correlation coefficients (R2) between the CO estimates using the 3 algorithms and COMRI were 0.271, 0.548 and 0.757, respectively. These R2 values were statistically significant and showed that COIPRAMcen performed best. The mean difference between the CO estimates using the 3 algorithms and COMRI were −0.15 ± 0.44, −0.07 ± 0.24 and −0.04 ± 0.17 L/min, respectively.

KW - Aortic pressure wave

KW - Cardiac output

KW - MRI

KW - Peripheral pressure wave

KW - PRAM

UR - https://www.scopus.com/pages/publications/105020257168

U2 - 10.1016/j.bspc.2025.108984

DO - 10.1016/j.bspc.2025.108984

M3 - Article

AN - SCOPUS:105020257168

SN - 1746-8094

VL - 113

JO - Biomedical Signal Processing and Control

JF - Biomedical Signal Processing and Control

IS - Part A

M1 - 108984

ER -

Estimation of cardiac output based on PRAM algorithm and ARX model from noninvasive radial pressure wave

Abstract

Bibliographical note

Funding

UN SDGs

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this