TY - JOUR
T1 - Early Detection of Human Epileptic Seizures Based on Intracortical Microelectrode Array Signals
AU - Park, Yun S.
AU - Cosgrove, G. Rees
AU - Madsen, Joseph R.
AU - Eskandar, Emad N.
AU - Hochberg, Leigh R.
AU - Cash, Sydney S.
AU - Truccolo, Wilson
N1 - Funding Information:
fairs B6453R (LRH); in part by the Doris Duke Charitable Foundation
Funding Information:
Manuscript received January 29, 2019; revised April 19, 2019; accepted May 23, 2019. Date of publication June 6, 2019; date of current version February 19, 2020. This work was supported in part by the National Institutes of Health–National Institute of Neurological Disorders and Stroke under Grants R01NS079533 (WT), K01NS057389 (WT), and R01NS062092 (SSC); in part by the Department of Veterans Affairs (Merit Review Award RX000668, WT), Office of Research and Development, Rehabilitation R&D Service, Department of Veterans Af-
Funding Information:
This work was supported in part by the National Institutes of Health-National Institute of Neurological Disorders and Stroke under Grants R01NS079533 (WT), K01NS057389 (WT), and R01NS062092 (SSC); in part by the Department of Veterans Affairs (Merit Review Award RX000668, WT), Office of Research and Development, Rehabilitation R&D Service, Department of Veterans Affairs B6453R (LRH); in part by the Doris Duke Charitable Foundation (LRH); in part by the Massachusetts General Hospital Deane Institute (LRH); and in part by a Postdoctoral Fellowship from the Epilepsy Foundation (YSP), and the Pablo J. Salame '88 Goldman Sachs endowed Assistant Professorship of Computational Neuroscience (WT).
Publisher Copyright:
© 1964-2012 IEEE.
PY - 2020/3
Y1 - 2020/3
N2 - Objective: We examine, for the first time, the use of intracortical microelectrode array (MEA) signals for early detection of human epileptic seizures. Methods: 4×4 mm2 96-channel-MEA recordings were obtained during neuro-monitoring preceding resective surgery in five participants. The participant-specific seizure-detection framework consisted of: first, feature extraction from local field potentials (LFPs) and multiunit activity (MUA); second, nonlinear cost-sensitive support vector machine (SVM) classification of ictal and interictal states based on LFP, MUA, and combined LFP-MUA (a SVM was trained for each participant separately); and third, Kalman filter postprocessing of SVM scoring functions. Performance was assessed on data including 17 seizures and 39.0 h interictal and preictal recordings. Results: The use of combined LFP-MUA features resulted in 100% sensitivity with short detection latency (average: 2.7 s; median: 2.5 s) and five false alarms (0.13/h). The average detection performance based on the area under the receiver operating characteristic corresponded to 0.97. Importantly, technically false alarms were related to epileptiform activity, subclinical seizures, and recording artifacts. Extreme gradient boosting classifiers ranked features based on LFP spectral coherence or MUA count among the top features for seizures characterized by spike-wave complexes, whereas features related to LFP power spectra were ranked higher for seizures characterized by sustained gamma LFP oscillations. Conclusion: The combination of intracortical LFP and MUA signals may allow reliable detection of human epileptic seizures by improving latency and false alarm rate. Significance: Intracortical MEAs provide promising signals for closed-loop seizure-control systems based on seizure early-detection in people with pharmacologically resistant epilepsies.
AB - Objective: We examine, for the first time, the use of intracortical microelectrode array (MEA) signals for early detection of human epileptic seizures. Methods: 4×4 mm2 96-channel-MEA recordings were obtained during neuro-monitoring preceding resective surgery in five participants. The participant-specific seizure-detection framework consisted of: first, feature extraction from local field potentials (LFPs) and multiunit activity (MUA); second, nonlinear cost-sensitive support vector machine (SVM) classification of ictal and interictal states based on LFP, MUA, and combined LFP-MUA (a SVM was trained for each participant separately); and third, Kalman filter postprocessing of SVM scoring functions. Performance was assessed on data including 17 seizures and 39.0 h interictal and preictal recordings. Results: The use of combined LFP-MUA features resulted in 100% sensitivity with short detection latency (average: 2.7 s; median: 2.5 s) and five false alarms (0.13/h). The average detection performance based on the area under the receiver operating characteristic corresponded to 0.97. Importantly, technically false alarms were related to epileptiform activity, subclinical seizures, and recording artifacts. Extreme gradient boosting classifiers ranked features based on LFP spectral coherence or MUA count among the top features for seizures characterized by spike-wave complexes, whereas features related to LFP power spectra were ranked higher for seizures characterized by sustained gamma LFP oscillations. Conclusion: The combination of intracortical LFP and MUA signals may allow reliable detection of human epileptic seizures by improving latency and false alarm rate. Significance: Intracortical MEAs provide promising signals for closed-loop seizure-control systems based on seizure early-detection in people with pharmacologically resistant epilepsies.
KW - Intracortical microelectrode array (MEA)
KW - human focal seizures
KW - multi-unit activity (MUA)
KW - seizure early-detection
UR - http://www.scopus.com/inward/record.url?scp=85080853191&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85080853191&partnerID=8YFLogxK
U2 - 10.1109/TBME.2019.2921448
DO - 10.1109/TBME.2019.2921448
M3 - Article
C2 - 31180831
AN - SCOPUS:85080853191
SN - 0018-9294
VL - 67
SP - 817
EP - 831
JO - IRE transactions on medical electronics
JF - IRE transactions on medical electronics
IS - 3
M1 - 8732415
ER -