Compartmental and Data-Based Modeling of Cerebral Hemodynamics

Nonlinear Analysis

Brandon Christian Henley, Dae C. Shin, Rong Zhang, Vasilis Z. Marmarelis

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Objective: As an extension to our study comparing a putative compartmental and data-based model of linear dynamic cerebral autoregulation (CA) and CO2-vasomotor reactivity (VR), we study the CA-VR process in a nonlinear context. Methods: We use the concept of principal dynamic modes (PDM) in order to obtain a compact and more easily interpretable input-output model. This in silico study permits the use of input data with a dynamic range large enough to simulate the classic homeostatic CA and VR curves using a putative structural model of the regulatory control of the cerebral circulation. The PDM model obtained using theoretical and experimental data are compared. Results: It was found that the PDM model was able to reflect accurately both the simulated static CA and VR curves in the associated nonlinear functions (ANFs). Similar to experimental observations, the PDM model essentially separates the pressure-flow relationship into a linear component with fast dynamics and nonlinear components with slow dynamics. In addition, we found good qualitative agreement between the PDMs representing the dynamic theoretical and experimental CO2-flow relationship. Conclusion: Under the modeling assumption and in light of other experimental findings, we hypothesize that PDMs obtained from experimental data correspond with passive fluid dynamical and active regulatory mechanisms. Significance: Both hypothesis-based and data-based modeling approaches can be combined to offer some insight into the physiological basis of PDM model obtained from human experimental data. The PDM modeling approach potentially offers a practical way to quantify the status of specific regulatory mechanisms in the CA-VR process.

Original languageEnglish (US)
Article number7508482
Pages (from-to)1078-1088
Number of pages11
JournalIEEE Transactions on Biomedical Engineering
Volume64
Issue number5
DOIs
StatePublished - May 1 2017

Fingerprint

Hemodynamics
Nonlinear analysis
Pulse width modulation

Keywords

  • Cerebral hemodynamics
  • nonparametric model
  • parametric model
  • principal dynamic modes (PDM)

ASJC Scopus subject areas

  • Biomedical Engineering

Cite this

Compartmental and Data-Based Modeling of Cerebral Hemodynamics : Nonlinear Analysis. / Henley, Brandon Christian; Shin, Dae C.; Zhang, Rong; Marmarelis, Vasilis Z.

In: IEEE Transactions on Biomedical Engineering, Vol. 64, No. 5, 7508482, 01.05.2017, p. 1078-1088.

Research output: Contribution to journalArticle

Henley, Brandon Christian ; Shin, Dae C. ; Zhang, Rong ; Marmarelis, Vasilis Z. / Compartmental and Data-Based Modeling of Cerebral Hemodynamics : Nonlinear Analysis. In: IEEE Transactions on Biomedical Engineering. 2017 ; Vol. 64, No. 5. pp. 1078-1088.
@article{8a300eac05ef49768550b91ca9d42ff2,
title = "Compartmental and Data-Based Modeling of Cerebral Hemodynamics: Nonlinear Analysis",
abstract = "Objective: As an extension to our study comparing a putative compartmental and data-based model of linear dynamic cerebral autoregulation (CA) and CO2-vasomotor reactivity (VR), we study the CA-VR process in a nonlinear context. Methods: We use the concept of principal dynamic modes (PDM) in order to obtain a compact and more easily interpretable input-output model. This in silico study permits the use of input data with a dynamic range large enough to simulate the classic homeostatic CA and VR curves using a putative structural model of the regulatory control of the cerebral circulation. The PDM model obtained using theoretical and experimental data are compared. Results: It was found that the PDM model was able to reflect accurately both the simulated static CA and VR curves in the associated nonlinear functions (ANFs). Similar to experimental observations, the PDM model essentially separates the pressure-flow relationship into a linear component with fast dynamics and nonlinear components with slow dynamics. In addition, we found good qualitative agreement between the PDMs representing the dynamic theoretical and experimental CO2-flow relationship. Conclusion: Under the modeling assumption and in light of other experimental findings, we hypothesize that PDMs obtained from experimental data correspond with passive fluid dynamical and active regulatory mechanisms. Significance: Both hypothesis-based and data-based modeling approaches can be combined to offer some insight into the physiological basis of PDM model obtained from human experimental data. The PDM modeling approach potentially offers a practical way to quantify the status of specific regulatory mechanisms in the CA-VR process.",
keywords = "Cerebral hemodynamics, nonparametric model, parametric model, principal dynamic modes (PDM)",
author = "Henley, {Brandon Christian} and Shin, {Dae C.} and Rong Zhang and Marmarelis, {Vasilis Z.}",
year = "2017",
month = "5",
day = "1",
doi = "10.1109/TBME.2016.2588438",
language = "English (US)",
volume = "64",
pages = "1078--1088",
journal = "IEEE Transactions on Biomedical Engineering",
issn = "0018-9294",
publisher = "IEEE Computer Society",
number = "5",

}

TY - JOUR

T1 - Compartmental and Data-Based Modeling of Cerebral Hemodynamics

T2 - Nonlinear Analysis

AU - Henley, Brandon Christian

AU - Shin, Dae C.

AU - Zhang, Rong

AU - Marmarelis, Vasilis Z.

PY - 2017/5/1

Y1 - 2017/5/1

N2 - Objective: As an extension to our study comparing a putative compartmental and data-based model of linear dynamic cerebral autoregulation (CA) and CO2-vasomotor reactivity (VR), we study the CA-VR process in a nonlinear context. Methods: We use the concept of principal dynamic modes (PDM) in order to obtain a compact and more easily interpretable input-output model. This in silico study permits the use of input data with a dynamic range large enough to simulate the classic homeostatic CA and VR curves using a putative structural model of the regulatory control of the cerebral circulation. The PDM model obtained using theoretical and experimental data are compared. Results: It was found that the PDM model was able to reflect accurately both the simulated static CA and VR curves in the associated nonlinear functions (ANFs). Similar to experimental observations, the PDM model essentially separates the pressure-flow relationship into a linear component with fast dynamics and nonlinear components with slow dynamics. In addition, we found good qualitative agreement between the PDMs representing the dynamic theoretical and experimental CO2-flow relationship. Conclusion: Under the modeling assumption and in light of other experimental findings, we hypothesize that PDMs obtained from experimental data correspond with passive fluid dynamical and active regulatory mechanisms. Significance: Both hypothesis-based and data-based modeling approaches can be combined to offer some insight into the physiological basis of PDM model obtained from human experimental data. The PDM modeling approach potentially offers a practical way to quantify the status of specific regulatory mechanisms in the CA-VR process.

AB - Objective: As an extension to our study comparing a putative compartmental and data-based model of linear dynamic cerebral autoregulation (CA) and CO2-vasomotor reactivity (VR), we study the CA-VR process in a nonlinear context. Methods: We use the concept of principal dynamic modes (PDM) in order to obtain a compact and more easily interpretable input-output model. This in silico study permits the use of input data with a dynamic range large enough to simulate the classic homeostatic CA and VR curves using a putative structural model of the regulatory control of the cerebral circulation. The PDM model obtained using theoretical and experimental data are compared. Results: It was found that the PDM model was able to reflect accurately both the simulated static CA and VR curves in the associated nonlinear functions (ANFs). Similar to experimental observations, the PDM model essentially separates the pressure-flow relationship into a linear component with fast dynamics and nonlinear components with slow dynamics. In addition, we found good qualitative agreement between the PDMs representing the dynamic theoretical and experimental CO2-flow relationship. Conclusion: Under the modeling assumption and in light of other experimental findings, we hypothesize that PDMs obtained from experimental data correspond with passive fluid dynamical and active regulatory mechanisms. Significance: Both hypothesis-based and data-based modeling approaches can be combined to offer some insight into the physiological basis of PDM model obtained from human experimental data. The PDM modeling approach potentially offers a practical way to quantify the status of specific regulatory mechanisms in the CA-VR process.

KW - Cerebral hemodynamics

KW - nonparametric model

KW - parametric model

KW - principal dynamic modes (PDM)

UR - http://www.scopus.com/inward/record.url?scp=85018949358&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85018949358&partnerID=8YFLogxK

U2 - 10.1109/TBME.2016.2588438

DO - 10.1109/TBME.2016.2588438

M3 - Article

VL - 64

SP - 1078

EP - 1088

JO - IEEE Transactions on Biomedical Engineering

JF - IEEE Transactions on Biomedical Engineering

SN - 0018-9294

IS - 5

M1 - 7508482

ER -