Abstracts

Rationale:
To evaluate the signal quality and noise of the Minder sub-scalp EEG system by comparing the activity recorded from the Minder sub-scalp EEG system with EEG signals acquired using standard scalp electrodes across multiple different time periods.




Methods:
Minder device and electrodes were implanted sub-scalp to enable chronic and continuous recordings. Recordings were evaluated through comparison of Minder EEG data with data acquired through two 7-day co-monitoring periods using international gold standard 10-20 scalp electrodes, performed at 4 and 24 weeks after Minder implantation. Recordings were evaluated for both subject-generated artefacts and spontaneous brain activity. Board-certified US electrophysiologists rated the scalp and sub scalp signals for their signal quality and noise and reviewed seizures, non-seizure electrographic activity and artefacts recorded during the co-monitoring period for comparison. In addition to co-monitoring, a further 6-month period was evaluated for seizures. Finally, a review of the last available month of data (within the 3-year follow up period) was also performed to evaluate ultra-long term recording stability. The study protocol was approved by St Vincent’s Hospital Melbourne IRB (HREC158/19) and registered under the CTN (ACTRN12619001587190).




Results:
During the co-monitoring period, Minder recorded any physiological (sleep spindles, vertex waves) or epileptiform (interictal discharges and seizures) signals that were also observed with the scalp electrodes. Minder was also able to differentiate resting state (eyes open/closed). Where seizures were detected using the gold standard 10-20 scalp system, all these seizures were also identified in the Minder EEG recordings. Overall signal quality ratings were performed by the reviewers and was found to be equivalent or near-equivalent to scalp EEG in >70% of recordings. Also during co-monitoring quality was very comparable between Minder and Scalp electrodes at 4 and 24 weeks (p=0.225). Signal noise was significantly lower in the Minder recordings for all EEG features during both co-monitoring periods (p=0.041). The Minder signal was stable over time and sensitivity was preserved for up to 3 years (study duration). In the cases where data was available after 6 months post implantation, EEG signals were detected in all cases and seizures could still be identified during the most recent month of recording.




Conclusions:
During the co-monitoring period the Minder system was able to record any signals that were also observed on the standard 10-20 scalp electrodes, thus showing Minder’s sensitivity to EEG signals is equivalent to scalp recordings. The Minder device can also record clinically significant EEG signals throughout the important range of frequencies, demonstrating the Minder system is an effective general purpose EEG recording device with equivalent signal quality and significantly lower signal noise when compared to standard scalp EEG. Furthermore, the Minder system permits continuous EEG recording with stable signal sensitivity over several years making it suitable for ultra-long term epilepsy monitoring.




Funding:
Funding was provided by Epiminder Pty. Ltd.