Single-trial reconstruction of finger-pinch forces from human motor-cortical activation measured by near-infrared spectroscopy (NIRS)

Isao Nambu, Rieko Osu, Masa aki Sato, Soichi Ando, Mitsuo Kawato, Eiichi Naito

Research output: Contribution to journalArticle

31 Citations (Scopus)

Abstract

Near-infrared spectroscopy (NIRS) has recently been used to measure human motor-cortical activation, enabling the classification of the content of a sensory-motor event such as whether the left or right hand was used. Here, we advance this NIRS application by demonstrating quantitative estimates of multiple sensory-motor events from single-trial NIRS signals. It is known that different degrees of sensory-motor activation are required to generate various hand/finger force levels. Thus, using a sparse linear regression method, we examined whether the temporal changes in different force levels could be reconstructed from NIRS signals. We measured the relative changes in oxyhemoglobin concentrations in the bilateral sensory-motor cortices while participants performed an isometric finger-pinch force production with their thumb and index finger by repeatedly exerting one of three target forces (25, 50, or 75% of the maximum voluntary contraction) for 12 s. To reconstruct the generated forces, we determined the regression parameters from the training datasets and applied these parameters to new test datasets to validate the parameters in the single-trial reconstruction. The temporal changes in the three different levels of generated forces, as well as the baseline resting state, could be reconstructed, even for the test datasets. The best reconstruction was achieved when using only the selected NIRS channels dominantly located in the contralateral sensory-motor cortex, and with a four second hemodynamic delay. These data demonstrate the potential for reconstructing different levels of external loads (forces) from those of the internal loads (activation) in the human brain using NIRS.

Original languageEnglish
Pages (from-to)628-637
Number of pages10
JournalNeuroImage
Volume47
Issue number2
DOIs
Publication statusPublished - 2009 Aug 15
Externally publishedYes

Fingerprint

Near-Infrared Spectroscopy
Fingers
Hand
Oxyhemoglobins
Thumb
Linear Models
Hemodynamics
Brain
Datasets

ASJC Scopus subject areas

  • Neurology
  • Cognitive Neuroscience

Cite this

Single-trial reconstruction of finger-pinch forces from human motor-cortical activation measured by near-infrared spectroscopy (NIRS). / Nambu, Isao; Osu, Rieko; Sato, Masa aki; Ando, Soichi; Kawato, Mitsuo; Naito, Eiichi.

In: NeuroImage, Vol. 47, No. 2, 15.08.2009, p. 628-637.

Research output: Contribution to journalArticle

Nambu, Isao ; Osu, Rieko ; Sato, Masa aki ; Ando, Soichi ; Kawato, Mitsuo ; Naito, Eiichi. / Single-trial reconstruction of finger-pinch forces from human motor-cortical activation measured by near-infrared spectroscopy (NIRS). In: NeuroImage. 2009 ; Vol. 47, No. 2. pp. 628-637.
@article{bc633702c8bf4a9286f416dc0738f97b,
title = "Single-trial reconstruction of finger-pinch forces from human motor-cortical activation measured by near-infrared spectroscopy (NIRS)",
abstract = "Near-infrared spectroscopy (NIRS) has recently been used to measure human motor-cortical activation, enabling the classification of the content of a sensory-motor event such as whether the left or right hand was used. Here, we advance this NIRS application by demonstrating quantitative estimates of multiple sensory-motor events from single-trial NIRS signals. It is known that different degrees of sensory-motor activation are required to generate various hand/finger force levels. Thus, using a sparse linear regression method, we examined whether the temporal changes in different force levels could be reconstructed from NIRS signals. We measured the relative changes in oxyhemoglobin concentrations in the bilateral sensory-motor cortices while participants performed an isometric finger-pinch force production with their thumb and index finger by repeatedly exerting one of three target forces (25, 50, or 75{\%} of the maximum voluntary contraction) for 12 s. To reconstruct the generated forces, we determined the regression parameters from the training datasets and applied these parameters to new test datasets to validate the parameters in the single-trial reconstruction. The temporal changes in the three different levels of generated forces, as well as the baseline resting state, could be reconstructed, even for the test datasets. The best reconstruction was achieved when using only the selected NIRS channels dominantly located in the contralateral sensory-motor cortex, and with a four second hemodynamic delay. These data demonstrate the potential for reconstructing different levels of external loads (forces) from those of the internal loads (activation) in the human brain using NIRS.",
author = "Isao Nambu and Rieko Osu and Sato, {Masa aki} and Soichi Ando and Mitsuo Kawato and Eiichi Naito",
year = "2009",
month = "8",
day = "15",
doi = "10.1016/j.neuroimage.2009.04.050",
language = "English",
volume = "47",
pages = "628--637",
journal = "NeuroImage",
issn = "1053-8119",
publisher = "Academic Press Inc.",
number = "2",

}

TY - JOUR

T1 - Single-trial reconstruction of finger-pinch forces from human motor-cortical activation measured by near-infrared spectroscopy (NIRS)

AU - Nambu, Isao

AU - Osu, Rieko

AU - Sato, Masa aki

AU - Ando, Soichi

AU - Kawato, Mitsuo

AU - Naito, Eiichi

PY - 2009/8/15

Y1 - 2009/8/15

N2 - Near-infrared spectroscopy (NIRS) has recently been used to measure human motor-cortical activation, enabling the classification of the content of a sensory-motor event such as whether the left or right hand was used. Here, we advance this NIRS application by demonstrating quantitative estimates of multiple sensory-motor events from single-trial NIRS signals. It is known that different degrees of sensory-motor activation are required to generate various hand/finger force levels. Thus, using a sparse linear regression method, we examined whether the temporal changes in different force levels could be reconstructed from NIRS signals. We measured the relative changes in oxyhemoglobin concentrations in the bilateral sensory-motor cortices while participants performed an isometric finger-pinch force production with their thumb and index finger by repeatedly exerting one of three target forces (25, 50, or 75% of the maximum voluntary contraction) for 12 s. To reconstruct the generated forces, we determined the regression parameters from the training datasets and applied these parameters to new test datasets to validate the parameters in the single-trial reconstruction. The temporal changes in the three different levels of generated forces, as well as the baseline resting state, could be reconstructed, even for the test datasets. The best reconstruction was achieved when using only the selected NIRS channels dominantly located in the contralateral sensory-motor cortex, and with a four second hemodynamic delay. These data demonstrate the potential for reconstructing different levels of external loads (forces) from those of the internal loads (activation) in the human brain using NIRS.

AB - Near-infrared spectroscopy (NIRS) has recently been used to measure human motor-cortical activation, enabling the classification of the content of a sensory-motor event such as whether the left or right hand was used. Here, we advance this NIRS application by demonstrating quantitative estimates of multiple sensory-motor events from single-trial NIRS signals. It is known that different degrees of sensory-motor activation are required to generate various hand/finger force levels. Thus, using a sparse linear regression method, we examined whether the temporal changes in different force levels could be reconstructed from NIRS signals. We measured the relative changes in oxyhemoglobin concentrations in the bilateral sensory-motor cortices while participants performed an isometric finger-pinch force production with their thumb and index finger by repeatedly exerting one of three target forces (25, 50, or 75% of the maximum voluntary contraction) for 12 s. To reconstruct the generated forces, we determined the regression parameters from the training datasets and applied these parameters to new test datasets to validate the parameters in the single-trial reconstruction. The temporal changes in the three different levels of generated forces, as well as the baseline resting state, could be reconstructed, even for the test datasets. The best reconstruction was achieved when using only the selected NIRS channels dominantly located in the contralateral sensory-motor cortex, and with a four second hemodynamic delay. These data demonstrate the potential for reconstructing different levels of external loads (forces) from those of the internal loads (activation) in the human brain using NIRS.

UR - http://www.scopus.com/inward/record.url?scp=67349172155&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=67349172155&partnerID=8YFLogxK

U2 - 10.1016/j.neuroimage.2009.04.050

DO - 10.1016/j.neuroimage.2009.04.050

M3 - Article

C2 - 19393320

AN - SCOPUS:67349172155

VL - 47

SP - 628

EP - 637

JO - NeuroImage

JF - NeuroImage

SN - 1053-8119

IS - 2

ER -