Computational tools for calculating alternative muscle force patterns during motion: A comparison of possible solutions

Saulo Martelli, Daniela Calvetti, Erkki Somersalo, Marco Viceconti, Fulvia Taddei

Research output: Contribution to journalArticle

Abstract

Comparing the available electromyography (EMG) and the related uncertainties with the space of muscle forces potentially driving the same motion can provide insights into understanding human motion in healthy and pathological neuromotor conditions. However, it is not clear how effective the available computational tools are in completely sample the possible muscle forces. In this study, we compared the effectiveness of Metabolica and the Null-Space algorithm at generating a comprehensive spectrum of possible muscle forces for a representative motion frame. The hip force peak during a selected walking trial was identified using a lower-limb musculoskeletal model. The joint moments, the muscle lever arms, and the muscle force constraints extracted from the model constituted the indeterminate equilibrium equation at the joints. Two spectra, each containing 200,000 muscle force samples, were calculated using Metabolica and the Null-Space algorithm. The full hip force range was calculated using optimization and compared with the hip force ranges derived from the Metabolica and the Null-Space spectra. The Metabolica spectrum spanned a much larger force range than the NS spectrum, reaching 811. N difference for the gluteus maximus intermediate bundle. The Metabolica hip force range exhibited a 0.3-0.4 BW error on the upper and lower boundaries of the full hip force range (3.4-11.3 BW), whereas the full range was imposed in the NS spectrum. The results suggest that Metabolica is well suited for exhaustively sample the spectrum of possible muscle recruitment strategy. Future studies will investigate the muscle force range in healthy and pathological neuromotor conditions.

Original languageEnglish
Pages (from-to)2097-2100
Number of pages4
JournalJournal of Biomechanics
Volume46
Issue number12
DOIs
Publication statusPublished - Aug 9 2013

Fingerprint

Muscle
Muscles
Hip
Joints
Electromyography
Walking
Uncertainty
Lower Extremity

Keywords

  • Bayesian muscle force simulator
  • Human locomotion
  • Level walking
  • Personalized hip loads
  • Statistical muscle recruitment
  • Stochastic muscle force generator
  • Subject-specific musculoskeletal models

ASJC Scopus subject areas

  • Orthopedics and Sports Medicine
  • Rehabilitation
  • Biophysics
  • Biomedical Engineering

Cite this

Computational tools for calculating alternative muscle force patterns during motion : A comparison of possible solutions. / Martelli, Saulo; Calvetti, Daniela; Somersalo, Erkki; Viceconti, Marco; Taddei, Fulvia.

In: Journal of Biomechanics, Vol. 46, No. 12, 09.08.2013, p. 2097-2100.

Research output: Contribution to journalArticle

@article{f79d8b27a2e54232a7367decd9931409,
title = "Computational tools for calculating alternative muscle force patterns during motion: A comparison of possible solutions",
abstract = "Comparing the available electromyography (EMG) and the related uncertainties with the space of muscle forces potentially driving the same motion can provide insights into understanding human motion in healthy and pathological neuromotor conditions. However, it is not clear how effective the available computational tools are in completely sample the possible muscle forces. In this study, we compared the effectiveness of Metabolica and the Null-Space algorithm at generating a comprehensive spectrum of possible muscle forces for a representative motion frame. The hip force peak during a selected walking trial was identified using a lower-limb musculoskeletal model. The joint moments, the muscle lever arms, and the muscle force constraints extracted from the model constituted the indeterminate equilibrium equation at the joints. Two spectra, each containing 200,000 muscle force samples, were calculated using Metabolica and the Null-Space algorithm. The full hip force range was calculated using optimization and compared with the hip force ranges derived from the Metabolica and the Null-Space spectra. The Metabolica spectrum spanned a much larger force range than the NS spectrum, reaching 811. N difference for the gluteus maximus intermediate bundle. The Metabolica hip force range exhibited a 0.3-0.4 BW error on the upper and lower boundaries of the full hip force range (3.4-11.3 BW), whereas the full range was imposed in the NS spectrum. The results suggest that Metabolica is well suited for exhaustively sample the spectrum of possible muscle recruitment strategy. Future studies will investigate the muscle force range in healthy and pathological neuromotor conditions.",
keywords = "Bayesian muscle force simulator, Human locomotion, Level walking, Personalized hip loads, Statistical muscle recruitment, Stochastic muscle force generator, Subject-specific musculoskeletal models",
author = "Saulo Martelli and Daniela Calvetti and Erkki Somersalo and Marco Viceconti and Fulvia Taddei",
year = "2013",
month = "8",
day = "9",
doi = "10.1016/j.jbiomech.2013.05.023",
language = "English",
volume = "46",
pages = "2097--2100",
journal = "Journal of Biomechanics",
issn = "0021-9290",
publisher = "Elsevier Limited",
number = "12",

}

TY - JOUR

T1 - Computational tools for calculating alternative muscle force patterns during motion

T2 - A comparison of possible solutions

AU - Martelli, Saulo

AU - Calvetti, Daniela

AU - Somersalo, Erkki

AU - Viceconti, Marco

AU - Taddei, Fulvia

PY - 2013/8/9

Y1 - 2013/8/9

N2 - Comparing the available electromyography (EMG) and the related uncertainties with the space of muscle forces potentially driving the same motion can provide insights into understanding human motion in healthy and pathological neuromotor conditions. However, it is not clear how effective the available computational tools are in completely sample the possible muscle forces. In this study, we compared the effectiveness of Metabolica and the Null-Space algorithm at generating a comprehensive spectrum of possible muscle forces for a representative motion frame. The hip force peak during a selected walking trial was identified using a lower-limb musculoskeletal model. The joint moments, the muscle lever arms, and the muscle force constraints extracted from the model constituted the indeterminate equilibrium equation at the joints. Two spectra, each containing 200,000 muscle force samples, were calculated using Metabolica and the Null-Space algorithm. The full hip force range was calculated using optimization and compared with the hip force ranges derived from the Metabolica and the Null-Space spectra. The Metabolica spectrum spanned a much larger force range than the NS spectrum, reaching 811. N difference for the gluteus maximus intermediate bundle. The Metabolica hip force range exhibited a 0.3-0.4 BW error on the upper and lower boundaries of the full hip force range (3.4-11.3 BW), whereas the full range was imposed in the NS spectrum. The results suggest that Metabolica is well suited for exhaustively sample the spectrum of possible muscle recruitment strategy. Future studies will investigate the muscle force range in healthy and pathological neuromotor conditions.

AB - Comparing the available electromyography (EMG) and the related uncertainties with the space of muscle forces potentially driving the same motion can provide insights into understanding human motion in healthy and pathological neuromotor conditions. However, it is not clear how effective the available computational tools are in completely sample the possible muscle forces. In this study, we compared the effectiveness of Metabolica and the Null-Space algorithm at generating a comprehensive spectrum of possible muscle forces for a representative motion frame. The hip force peak during a selected walking trial was identified using a lower-limb musculoskeletal model. The joint moments, the muscle lever arms, and the muscle force constraints extracted from the model constituted the indeterminate equilibrium equation at the joints. Two spectra, each containing 200,000 muscle force samples, were calculated using Metabolica and the Null-Space algorithm. The full hip force range was calculated using optimization and compared with the hip force ranges derived from the Metabolica and the Null-Space spectra. The Metabolica spectrum spanned a much larger force range than the NS spectrum, reaching 811. N difference for the gluteus maximus intermediate bundle. The Metabolica hip force range exhibited a 0.3-0.4 BW error on the upper and lower boundaries of the full hip force range (3.4-11.3 BW), whereas the full range was imposed in the NS spectrum. The results suggest that Metabolica is well suited for exhaustively sample the spectrum of possible muscle recruitment strategy. Future studies will investigate the muscle force range in healthy and pathological neuromotor conditions.

KW - Bayesian muscle force simulator

KW - Human locomotion

KW - Level walking

KW - Personalized hip loads

KW - Statistical muscle recruitment

KW - Stochastic muscle force generator

KW - Subject-specific musculoskeletal models

UR - http://www.scopus.com/inward/record.url?scp=84880698012&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84880698012&partnerID=8YFLogxK

U2 - 10.1016/j.jbiomech.2013.05.023

DO - 10.1016/j.jbiomech.2013.05.023

M3 - Article

C2 - 23791083

AN - SCOPUS:84880698012

VL - 46

SP - 2097

EP - 2100

JO - Journal of Biomechanics

JF - Journal of Biomechanics

SN - 0021-9290

IS - 12

ER -