Abstracts

Rationale: Dynamic network properties derived from normal interictal high density EEG recordings of patients with temporal lobe epilepsy (TLE) and normal control subjects are analyzed by applying dynamic mode decomposition (DMD). In comparison to previously utilized methods, DMD allows not only structural or functional connectivity between single electrodes or brain regions, but also provides information about dynamic network stability for a predefined frequency band. To investigate potential adaptive changes of neuronal networks during the course of drug refractory TLE, laterality indices (LI) of DMD are compared to the duration of disease. Methods: Three four minute epochs of normal resting state EEGs in 18 drug refractory TLE-patients (eight with onset within five years of the EEG (TLE_5), all with left sided seizure focus (L)), 10 with onset longer than 10 years ago (TLE_10), 6 with right sided (R), 4 with left sided (L) seizure focus) and 8 age-matched normal control subjects are analyzed. Laterality indices (LI) of DMD are calculated in each group in the frequency bands θ (4-8Hz), α (8-12 Hz), and β1 in (12-16 Hz). Results: With only few exceptions, LI’s of each epoch were intra-individually similar and did not differ between frequency bands. Seven of eight TLE_5 patients revealed a laterality associated with stronger coupling to the left (ipsilateral) hemisphere. TLE_10 patients presented either stronger ipsilateral (1/4 (L), 2/6 (R)), contralateral (0/4 (L), 1/6 (R)) or non-lateralized (3/4 (L), 3/6 (R)) coupling. LI’s of 1/8 control subjects indicated stronger left sided coupling, whereas 7/8 subjects revealed no or only minor laterality (to the right). Conclusions: Different LI’s depending on the duration of epilepsy in drug-refractory TLE may indicate reorganization of dynamic network properties, but so far little is known about physiological, pharmacological and pathological modifiers of DMD. This pilot study, however, may provide the basis for further longitudinal clinical studies investigating spatio-temporal dynamics in normal EEGs of epilepsy patients.  Funding: This work was in parts funded by the Forschungspreis 2015, Paderborn University, Germany.