### Abstract

Brain activation within and effective connectivity between two significantly activated left and right primary motor regions (M1 _{left} and M1 _{right}) were modelled using functional magnetic resonance imaging datasets of seven subjects obtained from an externally triggered unilateral and bilateral finger tapping experiment. Unilateral tapping exhibited contra lateral activation while bilateral tapping activated both the M1 _{left} and M1 _{right} regions. Unlike bilateral tapping that exhibited a relatively equal change of M1 _{left} and M1 _{right} signal at the coordinates of maximum intensity, the value is significantly lower in M1 _{left} as compared to M1 _{right} for unilateral tapping. The effective connectivity between M1 _{left} and M1 _{right} during bilateral tapping of hand fingers can be explained by a bilinear causal model, which is averagely preferred by five subjects (Dirichlet parameter estimate, α _{d} ≈ 5). The regional connectivities are however not gated (influenced) by any of the two M1s, ruling out the possibility of the non-linear behavior of connections between both M1s. This study has been able to fit the effective connectivity between M1left and M1 _{right} with a bilinear model that has the lowest free energy (F) of -644.70 and the largest likelihood (4.77 × 10 ^{39}), posterior (φ = 1.00), expected (r = 0.67) and exceedance (ψ = 0.92) probabilities.

Original language | English |
---|---|

Pages (from-to) | 178-212 |

Number of pages | 35 |

Journal | Malaysian Journal of Mathematical Sciences |

Volume | 6 |

Issue number | 2 |

Publication status | Published - Jul 2012 |

### Fingerprint

### Keywords

- Bayesian model selection
- Bilinear model
- Dynamic causal modeling
- Exceedance probability
- Expected probability

### ASJC Scopus subject areas

- Mathematics(all)

### Cite this

*Malaysian Journal of Mathematical Sciences*,

*6*(2), 178-212.

**Modelling activation and connectivity in the brain : An fMRI study during externally triggered finger tapping task.** / Yusoff, Ahmad Nazlim; Hamid, Khairiah Abdul.

Research output: Contribution to journal › Article

*Malaysian Journal of Mathematical Sciences*, vol. 6, no. 2, pp. 178-212.

}

TY - JOUR

T1 - Modelling activation and connectivity in the brain

T2 - An fMRI study during externally triggered finger tapping task

AU - Yusoff, Ahmad Nazlim

AU - Hamid, Khairiah Abdul

PY - 2012/7

Y1 - 2012/7

N2 - Brain activation within and effective connectivity between two significantly activated left and right primary motor regions (M1 left and M1 right) were modelled using functional magnetic resonance imaging datasets of seven subjects obtained from an externally triggered unilateral and bilateral finger tapping experiment. Unilateral tapping exhibited contra lateral activation while bilateral tapping activated both the M1 left and M1 right regions. Unlike bilateral tapping that exhibited a relatively equal change of M1 left and M1 right signal at the coordinates of maximum intensity, the value is significantly lower in M1 left as compared to M1 right for unilateral tapping. The effective connectivity between M1 left and M1 right during bilateral tapping of hand fingers can be explained by a bilinear causal model, which is averagely preferred by five subjects (Dirichlet parameter estimate, α d ≈ 5). The regional connectivities are however not gated (influenced) by any of the two M1s, ruling out the possibility of the non-linear behavior of connections between both M1s. This study has been able to fit the effective connectivity between M1left and M1 right with a bilinear model that has the lowest free energy (F) of -644.70 and the largest likelihood (4.77 × 10 39), posterior (φ = 1.00), expected (r = 0.67) and exceedance (ψ = 0.92) probabilities.

AB - Brain activation within and effective connectivity between two significantly activated left and right primary motor regions (M1 left and M1 right) were modelled using functional magnetic resonance imaging datasets of seven subjects obtained from an externally triggered unilateral and bilateral finger tapping experiment. Unilateral tapping exhibited contra lateral activation while bilateral tapping activated both the M1 left and M1 right regions. Unlike bilateral tapping that exhibited a relatively equal change of M1 left and M1 right signal at the coordinates of maximum intensity, the value is significantly lower in M1 left as compared to M1 right for unilateral tapping. The effective connectivity between M1 left and M1 right during bilateral tapping of hand fingers can be explained by a bilinear causal model, which is averagely preferred by five subjects (Dirichlet parameter estimate, α d ≈ 5). The regional connectivities are however not gated (influenced) by any of the two M1s, ruling out the possibility of the non-linear behavior of connections between both M1s. This study has been able to fit the effective connectivity between M1left and M1 right with a bilinear model that has the lowest free energy (F) of -644.70 and the largest likelihood (4.77 × 10 39), posterior (φ = 1.00), expected (r = 0.67) and exceedance (ψ = 0.92) probabilities.

KW - Bayesian model selection

KW - Bilinear model

KW - Dynamic causal modeling

KW - Exceedance probability

KW - Expected probability

UR - http://www.scopus.com/inward/record.url?scp=84867721664&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84867721664&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:84867721664

VL - 6

SP - 178

EP - 212

JO - Malaysian Journal of Mathematical Sciences

JF - Malaysian Journal of Mathematical Sciences

SN - 1823-8343

IS - 2

ER -