Seizure Prediction and Detection via Phase and Amplitude Lock Values
Author(s)
Myers, Mark H.; Padmanabha, Akshay; Hossain, Gahangir; de Jongh Curry, Amy L.; Blaha, Charles D.
DownloadMyers-2016-Seizure prediction.pdf (1.172Mb)
OPEN_ACCESS_POLICY
Open Access Policy
Creative Commons Attribution-Noncommercial-Share Alike
Terms of use
Metadata
Show full item recordAbstract
A robust seizure prediction methodology would enable a “closed-loop” system that would only activate as impending seizure activity is detected. Such a system would eliminate ongoing stimulation to the brain, thereby eliminating such side effects as coughing, hoarseness, voice alteration, and paresthesias (Murphy et al., 1998; Ben-Menachem, 2001), while preserving overall battery life of the system. The seizure prediction and detection algorithm uses Phase/Amplitude Lock Values (PLV/ALV) which calculate the difference of phase and amplitude between electroencephalogram (EEG) electrodes local and remote to the epileptic event. PLV is used as the seizure prediction marker and signifies the emergence of abnormal neuronal activations through local neuron populations. PLV/ALVs are used as seizure detection markers to demarcate the seizure event, or when the local seizure event has propagated throughout the brain turning into a grand-mal event. We verify the performance of this methodology against the “CHB-MIT Scalp EEG Database” which features seizure attributes for testing. Through this testing, we can demonstrate a high degree of sensivity and precision of our methodology between pre-ictal and ictal events.
Date issued
2016-03Department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer ScienceJournal
Frontiers in Human Neuroscience
Publisher
Frontiers Media S.A.
Citation
Myers, Mark H., Akshay Padmanabha, Gahangir Hossain, Amy L. de Jongh Curry, and Charles D. Blaha. "Seizure Prediction and Detection via Phase and Amplitude Lock Values." Front. Hum. Neurosci., 10:Article 80 (March 2016), p.1-9.
Version: Final published version
ISSN
1662-5161