Abstract
Objective
Seizure recurrence following surgery for temporal lobe (TL) epilepsy may be related to extratemporal epileptogenic foci, so‐called temporal‐plus (TL+) epilepsy. Here, we sought to leverage whole brain connectomic profiling in magnetoencephalography (MEG) to identify neural networks indicative of TL+ epilepsy in children.
Methods
Clinical and MEG data were analyzed for 121 children with TL and TL+ epilepsy spanning 20 years at the Hospital for Sick Children. Resting‐state connectomes were derived using the weighted phase lag index from neuromagnetic oscillations. Multidimensional associations between patient connectomes, TL versus TL+ epilepsy, seizure freedom, and clinical covariates were performed using a partial least squares (PLS) analysis. Bootstrap resampling statistics were performed to assess statistical significance.
Results
A single significant latent variable representing 66% of the variance in the data was identified with significant contributions from extent of epilepsy (TL vs TL+), duration of illness, and underlying etiology. This component was associated with significant bitemporal and frontotemporal connectivity in the theta, alpha, and beta bands. By extracting a brain score, representative of the observed connectivity profile, patients with TL epilepsy were dissociated from those with TL+, independent of their postoperative seizure outcome.
Significance
By analyzing 121 connectomes derived from MEG data using a PLS approach, we find that connectomic profiling could dissociate TL from TL+ epilepsy. These findings may inform patient selection for resective procedures and guide decisions surrounding invasive monitoring.
JUL