Epileptic networks Sample Clauses

Epileptic networks. In recent years interest has increased in investigating dynamic networks, (Xxxxxxxxxx et al., 2007). Two fundamental types of network investigated are (i) structural, which probe anatomical connectivity (white matter tracts) quantified by DTI and (ii) functional networks. In functional networks there is statistical coupling of dynamic activity between distinct regions, (Xxxxxx and Xxxx, 2012). Functional networks have been constructed using various methods including EEG, MEG and fMRI. There are also derivations of functional connectivity such as effective connectivity (Friston, 2011), which refers to the influence that one neural system exerts over another and uses a priori knowledge to assume directional influence; and xxxxxxx causality, which attempts to determine whether signals in a particular region can be predicted by preceding signals in a remote region. The definition of a network such as the brain is conceived as having two main components, nodes and edges, which can take several forms. Nodes may be in sensor space and represent a particular electrode in the case of EEG or may reflect a particular brain region in fMRI, with edges reflecting the strength of connections between nodes. Edges are defined as statistical associations between the time-series at the nodes; in functional networks these are inferred. In some instances this may be through investigation of node time series similarities. In the default mode network for example, fMRI regions during resting state are highly correlated and as such have strong edges. For the estimation of an association between nodes, a variety of different measures to assess statistical interdependencies can be chosen to define the degree of coupling between distinct regions. The most basic measure is correlation in the time domain, with a variety of options in the frequency domain explained in the following section in their relation to EEG networks in patients with epilepsy and healthy controls.