Rhythmogram-based analysis for continuous electrographic data of the human brain

Detection of short-lasting and ictal spike-and-wave discharges in around-the-ears EEG recordings in children with absence epilepsy

PURPOSE

Long-term ambulatory EEG recordings can improve the monitoring of absence epilepsy in children, but signal quality and increased review workload are a concern. We evaluated the feasibility of around-the-ears EEG arrays (cEEGrids) to capture 3-Hz short-lasting and ictal spike-and-wave discharges and assessed the performance of automated detection software in cEEGrids data. We compared patterns of bilateral synchronisation between short-lasting and ictal spike-and-wave discharges.

METHODS

We recruited children with suspected generalised epilepsy undergoing routine video-EEG monitoring and performed simultaneous cEEGrids recordings. We used ASSYST software to detect short-lasting 3-Hz spike-and-wave discharges (1-3 s) and ictal spike-and-wave discharges in the cEEGrids data. We assessed data quality and sensitivity of cEEGrids for spike-and-wave discharges in routine EEG. We determined the sensitivity and false detection rate for automated spike-and-wave discharge detection in cEEGrids data. We compared bihemispheric synchrony across the onset of short-lasting and ictal spike-and-wave discharges using the mean phase coherence in the 2-4 Hz frequency band.

RESULTS

We included nine children with absence epilepsy (median age = 11 y, range 8-15 y, nine females) and recorded 4 h and 27 min of cEEGrids data. The recordings from seven participants were suitable for quantitative analysis, containing 82 spike-and-wave discharges. The cEEGrids captured 58 % of all spike-and-wave discharges (median individual sensitivity: 100 %, range: 47-100 %). ASSYST detected 82 % of all spike-and-wave discharges (median: 100 %, range: 41-100 %) with a false detection rate of 48/h (median: 6/h, range: 0-154/h). The mean phase coherence significantly increased during short-lasting and ictal spike-and-wave discharges in the 500-ms pre-onset to 1-s post-onset interval.

CONCLUSIONS

cEEGrids are of variable quality for monitoring spike-and-wave discharges in children with absence epilepsy. ASSYST could facilitate the detection of short-lasting and ictal spike-and-wave discharges with clear periodic structures but with low specificity. A similar course of bihemispheric synchrony between short-lasting and ictal spike-and-wave discharges indicates that cortico-thalamic driving may be relevant for both types of spike-and-wave discharges.

An MEG compatible system for measuring skin conductance responses

We present the design of a low-cost system for recording galvanic skin conductance responses (SCRs) from humans in a magnetically shielded room (MSR) simultaneously to magnetoencephalography (MEG). Such a system was so far not available to the MEG community. Its availability is of utmost importance for neuroscience, since it will allow the concurrent assessment of the autonomic and central nervous system activity. The overall system design optimizes high signal to noise ratio (SNR) of SCRs and achieves minimal distortion of the MEG signal. Its development was based on a fiber-optic transformer, with voltage to optical transduction inside the MSR and demodulation outside the MSR. The system was calibrated and tested inside the MEG environment by using a 151-channel CTF whole head system (VSM MedTech Ltd.). MEG measurements were recorded simultaneously to SCRs from five healthy participants to test whether the developed system does not generate artifacts in the MEG data. Two measurements were performed for each participant; one without the system in the MSR, and one with the system in the MSR, connected to the participant and in operation. The data were analyzed using the time and frequency domains in separate statistical analysis. No significant differences were observed between the two sessions for any statistic index. Our results show that the system allows high quality simultaneous recordings of SCRs and MEG signals in the MSR, and can therefore be used as routine addendum to neuroscience experiments.