Cost function tuning improves muscle force estimation computed by static optimization during walking

V. Monaco, M. Coscia, S. Micera

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Muscle force estimation while a dynamic motor task is carried out still presents open questions. In particular, concerning locomotion, although the inverse dynamic based static optimization has been widely accepted as a suitable method to obtain reliable results, appropriate modifications of the object function may improve results. This paper was aimed at analyzing the sensitivity of estimated muscle forces when modifications of the objective function are adopted to better fit EMG signals of healthy subjects. A 7 links and 9 degrees of freedom biomechanical model accounting for 14 lower limb muscles, grouped in 9 equivalent actuators, was developed. Muscle forces were estimated by using the inverse dynamic based static optimization in which the performance criteria was the sum of muscle stresses raised to a certain n power. This exponent was gradually changed (from 2 to 100) and the agreement between force patterns and EMG signals was estimated by both the correlation coefficient and the Coactivation Index. Results suggested that force estimation can be improved by slightly modifying the cost function. In particular, with respect to adopted data, when the exponent belong to the interval between 2.75 and 4, estimated forces better captured general features of EMG signals. Concluding, a more reliable solution can be obtained by suitably tuning the cost function in order to fit EMG signals.

Original languageEnglish
Title of host publicationProceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS
Pages8263-8266
Number of pages4
DOIs
Publication statusPublished - 2011
Event33rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS 2011 - Boston, MA, United States
Duration: Aug 30 2011Sep 3 2011

Other

Other33rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS 2011
CountryUnited States
CityBoston, MA
Period8/30/119/3/11

Fingerprint

Cost functions
Walking
Muscle
Tuning
Costs and Cost Analysis
Muscles
Locomotion
Lower Extremity
Healthy Volunteers
Actuators

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition
  • Signal Processing
  • Biomedical Engineering
  • Health Informatics

Cite this

Monaco, V., Coscia, M., & Micera, S. (2011). Cost function tuning improves muscle force estimation computed by static optimization during walking. In Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS (pp. 8263-8266). [6092037] https://doi.org/10.1109/IEMBS.2011.6092037

Cost function tuning improves muscle force estimation computed by static optimization during walking. / Monaco, V.; Coscia, M.; Micera, S.

Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS. 2011. p. 8263-8266 6092037.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Monaco, V, Coscia, M & Micera, S 2011, Cost function tuning improves muscle force estimation computed by static optimization during walking. in Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS., 6092037, pp. 8263-8266, 33rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS 2011, Boston, MA, United States, 8/30/11. https://doi.org/10.1109/IEMBS.2011.6092037
Monaco V, Coscia M, Micera S. Cost function tuning improves muscle force estimation computed by static optimization during walking. In Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS. 2011. p. 8263-8266. 6092037 https://doi.org/10.1109/IEMBS.2011.6092037
Monaco, V. ; Coscia, M. ; Micera, S. / Cost function tuning improves muscle force estimation computed by static optimization during walking. Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS. 2011. pp. 8263-8266
@inproceedings{81abdb6da37b4e1982c30de83e3d07cc,
title = "Cost function tuning improves muscle force estimation computed by static optimization during walking",
abstract = "Muscle force estimation while a dynamic motor task is carried out still presents open questions. In particular, concerning locomotion, although the inverse dynamic based static optimization has been widely accepted as a suitable method to obtain reliable results, appropriate modifications of the object function may improve results. This paper was aimed at analyzing the sensitivity of estimated muscle forces when modifications of the objective function are adopted to better fit EMG signals of healthy subjects. A 7 links and 9 degrees of freedom biomechanical model accounting for 14 lower limb muscles, grouped in 9 equivalent actuators, was developed. Muscle forces were estimated by using the inverse dynamic based static optimization in which the performance criteria was the sum of muscle stresses raised to a certain n power. This exponent was gradually changed (from 2 to 100) and the agreement between force patterns and EMG signals was estimated by both the correlation coefficient and the Coactivation Index. Results suggested that force estimation can be improved by slightly modifying the cost function. In particular, with respect to adopted data, when the exponent belong to the interval between 2.75 and 4, estimated forces better captured general features of EMG signals. Concluding, a more reliable solution can be obtained by suitably tuning the cost function in order to fit EMG signals.",
author = "V. Monaco and M. Coscia and S. Micera",
year = "2011",
doi = "10.1109/IEMBS.2011.6092037",
language = "English",
isbn = "9781424441211",
pages = "8263--8266",
booktitle = "Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS",

}

TY - GEN

T1 - Cost function tuning improves muscle force estimation computed by static optimization during walking

AU - Monaco, V.

AU - Coscia, M.

AU - Micera, S.

PY - 2011

Y1 - 2011

N2 - Muscle force estimation while a dynamic motor task is carried out still presents open questions. In particular, concerning locomotion, although the inverse dynamic based static optimization has been widely accepted as a suitable method to obtain reliable results, appropriate modifications of the object function may improve results. This paper was aimed at analyzing the sensitivity of estimated muscle forces when modifications of the objective function are adopted to better fit EMG signals of healthy subjects. A 7 links and 9 degrees of freedom biomechanical model accounting for 14 lower limb muscles, grouped in 9 equivalent actuators, was developed. Muscle forces were estimated by using the inverse dynamic based static optimization in which the performance criteria was the sum of muscle stresses raised to a certain n power. This exponent was gradually changed (from 2 to 100) and the agreement between force patterns and EMG signals was estimated by both the correlation coefficient and the Coactivation Index. Results suggested that force estimation can be improved by slightly modifying the cost function. In particular, with respect to adopted data, when the exponent belong to the interval between 2.75 and 4, estimated forces better captured general features of EMG signals. Concluding, a more reliable solution can be obtained by suitably tuning the cost function in order to fit EMG signals.

AB - Muscle force estimation while a dynamic motor task is carried out still presents open questions. In particular, concerning locomotion, although the inverse dynamic based static optimization has been widely accepted as a suitable method to obtain reliable results, appropriate modifications of the object function may improve results. This paper was aimed at analyzing the sensitivity of estimated muscle forces when modifications of the objective function are adopted to better fit EMG signals of healthy subjects. A 7 links and 9 degrees of freedom biomechanical model accounting for 14 lower limb muscles, grouped in 9 equivalent actuators, was developed. Muscle forces were estimated by using the inverse dynamic based static optimization in which the performance criteria was the sum of muscle stresses raised to a certain n power. This exponent was gradually changed (from 2 to 100) and the agreement between force patterns and EMG signals was estimated by both the correlation coefficient and the Coactivation Index. Results suggested that force estimation can be improved by slightly modifying the cost function. In particular, with respect to adopted data, when the exponent belong to the interval between 2.75 and 4, estimated forces better captured general features of EMG signals. Concluding, a more reliable solution can be obtained by suitably tuning the cost function in order to fit EMG signals.

UR - http://www.scopus.com/inward/record.url?scp=84055213018&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84055213018&partnerID=8YFLogxK

U2 - 10.1109/IEMBS.2011.6092037

DO - 10.1109/IEMBS.2011.6092037

M3 - Conference contribution

SN - 9781424441211

SP - 8263

EP - 8266

BT - Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS

ER -