Abstracts

Rationale: A self-reported epileptic seizure diary is the current standard in clinical practice to guide anti-seizure treatment despite its known limitations. Ultra long-term monitoring with subcutaneous EEG (sqEEG) offers a novel and objective alternative to record electrographic seizures in an outpatient setting. This modality provides extensive data that needs to be reduced through an algorithm-based automatic seizure detection to indicate periods of potential seizure activity to make it feasible for physicians to review the seizure burden.

Methods: Two cohorts of subjects using sqEEG were analyzed, including nine PWE and 12 healthy subjects, recording a total of 965 days. The automatic seizure detections of a deep-neural-network algorithm were compared to annotations from three human experts. Moreover, the objective sqEEG-based seizure counts were compared against the self-reported diaries.

Results: Based on 94 seizures the overall sensitivity was 86 % (median 80 %, range 69 - 100 % when PWE were evaluated individually), and the median false detection rate was 2.4 detections per 24 hours (range: 2.0 - 13.0). Data reduction ratios were 99.6 % in PWE and 99.9 % in the control group. In comparison with self-reported seizure diaries, six out of eight PWE were considered examples of clinical cases where the objective sqEEG-based seizure monitoring would provide valuable insights to optimize the epilepsy treatment strategy.

Conclusions: We have demonstrated that sqEEG-based semi-automatic seizure detection/review process can be performed with high sensitivity and clinically applicable specificity. These findings highlight the potential of sqEEG-based outpatient seizure monitoring as an objective alternative to self-reported seizure diaries. This could be a pathway for treatment optimization in patients with self-reporting challenges.

Funding: None