Authors :
Presenting Author: Petroula Laiou, PhD – King's College London
Zeljko Kraljevic, PhD – King's College London; Antonio Valentin, MD – King's College London; Pedro Viana, MD – King's College London; Chirag Mehra, MD – King's College London; Richard Dobson, PhD – King's College London; Andreas Schulze-Bonhage, MD – University of Freiburg; Matthias Duempelmann, PhD – University of Freiburg; Timothy Denison, PhD – University of Oxford; Joel Winston, MD – King's College London; Mark Richardson, MD – King's College London
Rationale:
Seizure unpredictability is considered the most impactful aspect in the lives of people with epilepsy. Previously published approaches to seizure forecasting, analyzed intracranial EEG recordings (iEEG) showed that seizures can be forecasted above chance levels. Although passive observation of the brain might provide some insights, active perturbation of the cortex and measuring the cortical response may provide much more direct information about cortical excitability. Here we investigate whether seizures can be forecasted by stimulating the cortex via intracranial electrodes and measuring cortical response from the iEEG.
Methods:
We analyze eight patients who were admitted to King’s College Hospital for presurgical intracranial implantation. During their stay, they underwent prolonged single pulse electrical stimulation for 19 hours (mean; range 14-25 hours). Stimuli were delivered every five minutes to a constant pair of electrodes. All patients experienced at least one clinical seizure. We extracted quantitative features from the iEEG post-stimulus response and developed a logistic regression algorithm to estimate the seizure likelihood at each stimulus. To evaluate the algorithm’s performance, we used Improvement over chance, sensitivity, time spent in warning and Brier Skill score.
Results:
In seven out of eight patients, seizures could be forecasted above chance levels. The average Improvement over chance was 0.74 (standard deviation: 0.3), average sensitivity was 0.88 (standard deviation: 0.35), average time spent in warning was 0.14 (standard deviation: 0.09), average Brier Skill Score was 0.33 (standard deviation: 0.22), and average forecasting horizon was 73.86min (standard deviation: 36.41min).
Conclusions:
These results suggest that cortical response to electrical stimulation may aid in the development of seizure forecasting algorithms as well as in the design of novel implantable devices that deliver electrical stimulation to control seizures.
Funding:
This work was supported by the MRC IAA 2021 Kings College London (MR/X502923/1).