ABSTRACT
OBJECTIVE: To characterize generalized spike-wave paroxysms (GSW) in children with generalized genetic epilepsy (GGE). METHODS: We annotated 15-19 channel scalp EEGs from a retrospective cohort from patients with a variety of GGE syndromes. Connectivity, entropy, frequency, power, spike-amplitudes were compared with a normal baseline activity and analyzed for the effect of age and sex. Cluster analysis was used to group spike-topographies between patients. RESULTS: In total, 864 GSWs from 100 patients aged 2-18 were analyzed. Age had a significant effect on peak frequency, entropy and connectivity. Female sex was associated with significantly higher probability of positive responsiveness to photic stimulation (OR 4.28, CI [1.65, 11.73], p = 0.0036). Entropy decreases significantly during GSW (D = -0.29, CI [-0.31, -0.27], p ⪠0.0001) and connectivity significantly increases (D = 0.39, CI [0.36, 0.40], p ⪠0.0001). Within patient spike-voltage maps exhibit remarkable consistency between spikes. Spike-topographies cluster together to predict age, connectivity and entropy. CONCLUSIONS: A quantitative characterization is possible and reveals significant relationships between age, sex and spike characteristics and multidimensional EEG features. SIGNIFICANCE: Quantitative GSW characterization can capture aspects from traditional qualitative GSW analysis while being unaffected by intra- and interrater variation and this may be useful for multidimensional predictors of patient outcomes in GGE in the future.
