An interictal EEG spectral metric for temporal lobe epilepsy lateralization

Abstract

Objective Visually-obvious abnormalities in the resting baseline EEG – slowing, spiking and high-frequency oscillations (HFOs) - are cardinal, though incompletely understood, features of the seizure onset zone in focal epilepsy. We hypothesized that evidence of cortical network dysfunction in temporal lobe epilepsy (TLE) would persist in the absence of visually-classifiable abnormalities in the baseline EEG recorded within the conventional passband, and that metrics of such dysfunction could serve as a lateralizing diagnostic in TLE. Methods Epochs of resting EEG without significant abnormalities in light sleep over several days were compared between a group of 10 patients with proven TLE and 10 subjects without epilepsy. A novel laterality metric computed from the line length of normalized power spectra from the temporal channels was compared between the two groups. Results Significant group differences in spectral line length laterality metric were found between the TLE and control group. At the individual level, seven of 10 TLE patients had highly significant laterality metrics, all concordant with the known laterality of their disease. Significance Detailed spectral analysis offers novel insight into TLE network behavior, independent of the orthodox abnormalities of EEG slowing, spikes or HFOs. The results may be deployed in a practical diagnostic manner, offer insight into the EEG manifestations of disordered cellular network architecture in TLE, and maybe understood through simple analogy with the theory of linear time-invariant physical systems.

Regional Connectivity Changes According to Seizure Outcome of Temporal Epilepsy Surgery: A Magnetoencephalography Study

“…Although we did not observe difference of spectral power between the two groups that reach statistical significance, a tendency for decrease of delta power in areas (ROI 6) in seizurefree patients compared to ones with persistent seizures. A previous study on EEG free from visual abnormality showed difference of spectrum-based metrics between TLE and healthy subjects [19]. Because interictal lesional slowing is related with epileptic network in TLE patients [20], our data suggest that still remaining abnormal network is related with recurrence of seizures.…”

A unified statistical model for the human electrocorticogram

“…We have demonstrated previously (Kalamangalam et al, 2014a; Kalamangalam et al, 2014b; Kalamangalam and Slater, 2015) that such a formalism for the EEG and ECoG – a continuum hypothesis, though not necessarily restricted to a particular distribution – naturally lends itself to the emergence of large-scale synchronous (i.e., epileptic) activity. Synchronization involves the oscillators ‘condensing’ to a smaller distribution by increasing their mutual coupling, leading to the appearance of large amplitude coherent rhythms in the time domain as the coupling parameters cross a threshold.…”

“…The log-normal model is a continuum approximation in frequency space: the ECoG is viewed as a collection of oscillators whose frequency distributions (via a combination of intrinsic frequencies and other frequencies arising from self-interaction) approximate that statistical distribution. We have demonstrated previously (Kalamangalam et al, 2014a; Kalamangalam et al, 2014b; Kalamangalam and Slater, 2015) that such a formalism for the EEG and ECoG – a continuum hypothesis, though not necessarily restricted to a particular distribution – naturally lends itself to the emergence of large-scale synchronous (i.e., epileptic) activity. Synchronization involves the oscillators ‘condensing’ to a smaller distribution by increasing their mutual coupling, leading to the appearance of large amplitude coherent rhythms in the time domain as the coupling parameters cross a threshold.…”

“…The model also provides a conceptual bridge to understanding the transition from baseline EEG/ECoG to several epilepsy-related phenomena described in our previous work - the background rhythms of temporal lobe epilepsy (Kalamangalam et al, 2014a), afterdischarges (Kalamangalam et al, 2014b), and periodic lateralized epileptiform discharges (Kalamangalam and Slater, 2015). …”

“…In this work we use the high fidelity clinical datasets available to us from long term recordings in epilepsy patients undergoing intracerebral EEG (i.e., ECoG) recordings to propose a novel, unified, and parsimonious statistical description of the resting ECoG that synthesizes and reconciles a number of previous observations. The model also provides a conceptual bridge to understanding the transition from baseline EEG/ECoG to several epilepsy-related phenomena described in our previous work - the background rhythms of temporal lobe epilepsy (Kalamangalam et al, 2014a), afterdischarges (Kalamangalam et al, 2014b), and periodic lateralized epileptiform discharges (Kalamangalam and Slater, 2015).…”