Dynamic Changes in Brain Activations and Functional Connectivity during Affectively Different Tactile Stimuli

Abstract

In the present study, we compared brain activations produced by pleasant, neutral and unpleasant touch, to the anterior lateral surface of lower leg of human subjects. It was found that several brain regions, including the contralateral primary somatosensory area (SI), bilateral secondary somatosensory area (SII), as well as contralateral middle and posterior insula cortex were commonly activated under the three touch conditions. In addition, pleasant and unpleasant touch conditions shared a few brain regions including the contralateral posterior parietal cortex (PPC) and bilateral premotor cortex (PMC). Unpleasant touch specifically activated a set of pain-related brain regions such as contralateral supplementary motor area (SMA) and dorsal parts of bilateral anterior cingulated cortex, etc. Brain regions specifically activated by pleasant touch comprised bilateral lateral orbitofrontal cortex (OFC), posterior cingulate cortex (PCC), medial prefrontal cortex (mPFC), intraparietal cortex and left dorsal lateral prefrontal cortex (DLPFC). Using a novel functional connectivity model based on graph theory, we showed that a series of brain regions related to affectively different touch had significant functional connectivity during the resting state. Furthermore, it was found that such a network can be modulated between affectively different touch conditions.

Appendix

In this study, functional connectivity network analysis is based on graph theory. Therefore, the nodes denote the brain regions (ROIs) and the links denote the connections or information flow. In this way, we can define the total connectivity degree $\Upgamma _{i}$ of node i in a graph as the sum of all the connectivity degrees between i and all other nodes, i.e.,

$$ \Upgamma _{i} = {\sum\nolimits_{j = 1}^n {\eta _{{ij}} } } $$

where $\eta _{{ij}}$ is the connectivity degree between the node i and the node \, defined by the exponential function of the distance between them (Lopez and Sanjuan 2002),

$$ \eta _{{ij}} = e^{{ - \xi d_{{ij}} }} $$

where, ξ is a real positive constant, measuring how the strength of the relationship decreases with the distance between the two nodes [ξ is a subjective selection and discussed by Lopez and Sanjuan (2002) and is here fixed to ξ = 2], and d _ij is the distance between the two nodes, calculated as a hyperbolic correlation measure (Golay et al. 1998),

$$ d_{{ij}} = (1 - c_{{ij}} )/(1 + c_{{ij}} ) $$

where c _ij represents the Pearson correlation coefficient between the two nodes (i.e., cross-correlating two mean time series of the above).

In addition, as there are different touch conditions and different preprocessing in this study, we normalized $\Upgamma _{i}$ of a node i, namely,

$$ \ifmmode\expandafter\bar\else\expandafter\=\fi{\Upgamma }_{i} = \Upgamma _{i} /{\sum\nolimits_{j = 1}^n {\Upgamma _{j} } } $$

The normality of the distribution of the $\ifmmode\expandafter\bar\else\expandafter\=\fi{\Upgamma }_{i}$ values for all conditions was tested using the Jarque-Bera test.