King's College London

Research portal

Seizure forecasting using minimally invasive, ultra-long-term subcutaneous EEG: Generalizable cross-patient models

Research output: Contribution to journalArticlepeer-review

Tal Pal Attia, Pedro F. Viana, Mona Nasseri, Jonas Duun-Henriksen, Andrea Biondi, Joel S. Winston, Isabel P. Martins, Ewan S. Nurse, Matthias Dümpelmann, Gregory A. Worrell, Andreas Schulze-Bonhage, Dean R. Freestone, Troels W. Kjaer, Benjamin H. Brinkmann, Mark P. Richardson

Original languageEnglish
JournalEpilepsia
DOIs
Accepted/In press2022

Bibliographical note

Funding Information: This work was funded by the “My Seizure Gauge” grant provided by the Epilepsy Innovation Institute, a research program of the Epilepsy Foundation of America. MPR is supported by the National Institute for Health and Care Research Biomedical Research Centre at the South London and Maudsley National Health Service (NHS) Foundation Trust; the Medical Research Council Centre for Neurodevelopmental Disorders (MR/ N026063/1); and the Remote Assessment of Disease and Relapse ‐ Central Nervous System (RADAR‐CNS project funded by the European Commission ( www.radar‐cns.org , grant agreement 115902). BHB is supported by the Mayo Clinic Neurology AI program and by the National Institutes of Health (UG3 NS123066). We would like to thank the Neurosurgical team (Mr. Harishchandra Srinivasan, Mr. Harutomo Hasegawa, and Mr. Richard Selway) involved in the implantation procedures at King's College Hospital NHS Foundation Trust. Ultimately, we would like to acknowledge all patients who participated in this study. Funding Information: JDH is an employee of UNEEG medical A/S. ESN and DF are employees and shareholders of Seer Medical. BHB has equity in Cadence Neurosciences, has research funding from Seer Medical, and has received research devices from Medtronic Inc. at no cost. MPR has been a member of ad hoc advisory boards for UNEEG medical A/S. PFV received a payment from UNEEG medical A/S for data annotation in an unrelated research study. TWK consults for UNEEG medical A/S. No other authors have conflicts to declare. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines. Publisher Copyright: © 2022 International League Against Epilepsy.

King's Authors

Abstract

This study describes a generalized cross-patient seizure-forecasting approach using recurrent neural networks with ultra-long-term subcutaneous EEG (sqEEG) recordings. Data from six patients diagnosed with refractory epilepsy and monitored with an sqEEG device were used to develop a generalized algorithm for seizure forecasting using long short-term memory (LSTM) deep-learning classifiers. Electrographic seizures were identified by a board-certified epileptologist. One-minute data segments were labeled as preictal or interictal based on their relationship to confirmed seizures. Data were separated into training and testing data sets, and to compensate for the unbalanced data ratio in training, noise-added copies of preictal data segments were generated to expand the training data set. The mean and standard deviation (SD) of the training data were used to normalize all data, preserving the pseudo-prospective nature of the analysis. Different architecture classifiers were trained and tested using a leave-one-patient-out cross-validation method, and the area under the receiver-operating characteristic (ROC) curve (AUC) was used to evaluate the performance classifiers. The importance of each input signal was evaluated using a leave-one-signal-out method with repeated training and testing for each classifier. Cross-patient classifiers achieved performance significantly better than chance in four of the six patients and an overall mean AUC of 0.602 ± 0.126 (mean ± SD). A time in warning of 37.386% ± 5.006% (mean ± std) and sensitivity of 0.691 ± 0.068 (mean ± std) were observed for patients with better than chance results. Analysis of input channels showed a significant contribution (p <.05) by the Fourier transform of signals channels to overall classifier performance. The relative contribution of input signals varied among patients and architectures, suggesting that the inclusion of all signals contributes to robustness in a cross-patient classifier. These early results show that it is possible to forecast seizures training with data from different patients using two-channel ultra-long-term sqEEG.

View graph of relations

© 2020 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454