Magnetoencephalography Imaging Reveals Abnormal Information Flow in Temporal Lobe Epilepsy.

Abstract

Background/Introduction:Widespread network disruption has been hypothesized to be an important predictor of outcomes in patients with refractory temporal lobe epilepsy (TLE). Most studies examining functional network disruption in epilepsy have largely focused on the symmetric bidirectional metrics of the strength of network connections. However, a more complete description of network dysfunction impacts in epilepsy requires an investigation of the potentially more sensitive directional metrics of information flow.Methods:This study describes a whole-brain magnetoencephalography-imaging approach to examine resting-state directional information flow networks, quantified by phase-transfer entropy (PTE), in patients with TLE compared with healthy controls (HCs). Associations between PTE and clinical characteristics of epilepsy syndrome are also investigated.Results:Deficits of information flow were specific to alpha-band frequencies. In alpha band, while HCs exhibit a clear posterior-to-anterior directionality of information flow, in patients with TLE, this pattern of regional information outflow and inflow was significantly altered in the frontal and occipital regions. The changes in information flow within the alpha band in selected brain regions were correlated with interictal spike frequency and duration of epilepsy.Conclusions:Impaired information flow is an important dimension of network dysfunction associated with the pathophysiological mechanisms of TLE.Impact statementA complete description of network dysfunction in temporal lobe epilepsy (TLE) requires investigation of the flow of information in brain networks. In this whole-brain magnetoencephalography imaging study, using an information theory measure, phase-transfer entropy, we examine the strength and directionality of information flow in patients with TLE. Abnormal information flow was documented and was associated with clinical measures related to epilepsy in certain brain regions. The approach described here may provide a useful framework to investigate network dysfunction in other neurological and psychiatric disorders.