Huxley-type cross-bridge models in largeish-scale musculoskeletal models; an evaluation of computational cost

Research output: Contribution to JournalArticleAcademicpeer-review

Abstract

A Huxley-type cross-bridge model is attractive because it is inspired by our current understanding of the processes underlying muscle contraction, and because it provides a unified description of muscle's mechanical behavior and metabolic energy expenditure. In this study, we determined the computational cost for task optimization of a largeish-scale musculoskeletal model in which muscles are represented by a 2-state Huxley-type cross-bridge model. Parameter values defining the rate functions of the Huxley-type cross-bridge model could be chosen such that the steady-state force-velocity relation resembled that of a Hill-type model. Using these parameter values, maximum-height squat jumping was used as the example task to evaluate the computational cost of task optimization for a skeletal model driven by a Huxley-type cross-bridge model. The optimal solutions for the Huxley- and Hill-type muscle models were similar for all mechanical variables considered. Computational cost of the Huxley-type cross-bridge model was much higher than that of the Hill-type model. Compared to the Hill-type model, the number of state variables per muscle was large (2 vs about 18,000), the integration step size had to be about 100 times smaller, and the computational cost per integration step was about 100 times higher.

Original languageEnglish
Pages (from-to)43-48
Number of pages6
JournalJournal of Biomechanics
Volume83
Early online date22 Nov 2018
DOIs
Publication statusPublished - 23 Jan 2019

Fingerprint

Costs and Cost Analysis
Muscles
Costs
Muscle
Muscle Contraction
Energy Metabolism
Mechanical properties

Keywords

  • Forward dynamics
  • Hill
  • Huxley
  • Muscle model
  • Optimization
  • Vertical jumping

Cite this

@article{0e08f56eba654a6986d86faee4d7456b,
title = "Huxley-type cross-bridge models in largeish-scale musculoskeletal models; an evaluation of computational cost",
abstract = "A Huxley-type cross-bridge model is attractive because it is inspired by our current understanding of the processes underlying muscle contraction, and because it provides a unified description of muscle's mechanical behavior and metabolic energy expenditure. In this study, we determined the computational cost for task optimization of a largeish-scale musculoskeletal model in which muscles are represented by a 2-state Huxley-type cross-bridge model. Parameter values defining the rate functions of the Huxley-type cross-bridge model could be chosen such that the steady-state force-velocity relation resembled that of a Hill-type model. Using these parameter values, maximum-height squat jumping was used as the example task to evaluate the computational cost of task optimization for a skeletal model driven by a Huxley-type cross-bridge model. The optimal solutions for the Huxley- and Hill-type muscle models were similar for all mechanical variables considered. Computational cost of the Huxley-type cross-bridge model was much higher than that of the Hill-type model. Compared to the Hill-type model, the number of state variables per muscle was large (2 vs about 18,000), the integration step size had to be about 100 times smaller, and the computational cost per integration step was about 100 times higher.",
keywords = "Forward dynamics, Hill, Huxley, Muscle model, Optimization, Vertical jumping",
author = "{van Soest}, {A. J.“Knoek”} and Casius, {L. J.R.} and Lemaire, {K. K.}",
year = "2019",
month = "1",
day = "23",
doi = "10.1016/j.jbiomech.2018.11.021",
language = "English",
volume = "83",
pages = "43--48",
journal = "Journal of Biomechanics",
issn = "0021-9290",
publisher = "Elsevier Limited",

}

TY - JOUR

T1 - Huxley-type cross-bridge models in largeish-scale musculoskeletal models; an evaluation of computational cost

AU - van Soest, A. J.“Knoek”

AU - Casius, L. J.R.

AU - Lemaire, K. K.

PY - 2019/1/23

Y1 - 2019/1/23

N2 - A Huxley-type cross-bridge model is attractive because it is inspired by our current understanding of the processes underlying muscle contraction, and because it provides a unified description of muscle's mechanical behavior and metabolic energy expenditure. In this study, we determined the computational cost for task optimization of a largeish-scale musculoskeletal model in which muscles are represented by a 2-state Huxley-type cross-bridge model. Parameter values defining the rate functions of the Huxley-type cross-bridge model could be chosen such that the steady-state force-velocity relation resembled that of a Hill-type model. Using these parameter values, maximum-height squat jumping was used as the example task to evaluate the computational cost of task optimization for a skeletal model driven by a Huxley-type cross-bridge model. The optimal solutions for the Huxley- and Hill-type muscle models were similar for all mechanical variables considered. Computational cost of the Huxley-type cross-bridge model was much higher than that of the Hill-type model. Compared to the Hill-type model, the number of state variables per muscle was large (2 vs about 18,000), the integration step size had to be about 100 times smaller, and the computational cost per integration step was about 100 times higher.

AB - A Huxley-type cross-bridge model is attractive because it is inspired by our current understanding of the processes underlying muscle contraction, and because it provides a unified description of muscle's mechanical behavior and metabolic energy expenditure. In this study, we determined the computational cost for task optimization of a largeish-scale musculoskeletal model in which muscles are represented by a 2-state Huxley-type cross-bridge model. Parameter values defining the rate functions of the Huxley-type cross-bridge model could be chosen such that the steady-state force-velocity relation resembled that of a Hill-type model. Using these parameter values, maximum-height squat jumping was used as the example task to evaluate the computational cost of task optimization for a skeletal model driven by a Huxley-type cross-bridge model. The optimal solutions for the Huxley- and Hill-type muscle models were similar for all mechanical variables considered. Computational cost of the Huxley-type cross-bridge model was much higher than that of the Hill-type model. Compared to the Hill-type model, the number of state variables per muscle was large (2 vs about 18,000), the integration step size had to be about 100 times smaller, and the computational cost per integration step was about 100 times higher.

KW - Forward dynamics

KW - Hill

KW - Huxley

KW - Muscle model

KW - Optimization

KW - Vertical jumping

UR - http://www.scopus.com/inward/record.url?scp=85058234363&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85058234363&partnerID=8YFLogxK

U2 - 10.1016/j.jbiomech.2018.11.021

DO - 10.1016/j.jbiomech.2018.11.021

M3 - Article

VL - 83

SP - 43

EP - 48

JO - Journal of Biomechanics

JF - Journal of Biomechanics

SN - 0021-9290

ER -