Piotr Mirowski1, Deepak Madhavan2, Yann LeCun3, Ruben Kuzniecky4. 1. Courant Institute of Mathematical Sciences, New York University, 719 Broadway, New York, NY 10003, USA. Electronic address: piotr.mirowski@computer.org. 2. Department of Neurological Sciences, 982045 University of Nebraska Medical Center, Omaha, NE 68198, USA. 3. Courant Institute of Mathematical Sciences, New York University, 719 Broadway, New York, NY 10003, USA. 4. New York University Comprehensive Epilepsy Center, 223 East 34th St., New York, NY 10016, USA.
Abstract
OBJECTIVE: Research in seizure prediction from intracranial EEG has highlighted the usefulness of bivariate measures of brainwave synchronization. Spatio-temporal bivariate features are very high-dimensional and cannot be analyzed with conventional statistical methods. Hence, we propose state-of-the-art machine learning methods that handle high-dimensional inputs. METHODS: We computed bivariate features of EEG synchronization (cross-correlation, nonlinear interdependence, dynamical entrainment or wavelet synchrony) on the 21-patient Freiburg dataset. Features from all channel pairs and frequencies were aggregated over consecutive time points, to form patterns. Patient-specific machine learning-based classifiers (support vector machines, logistic regression or convolutional neural networks) were trained to discriminate interictal from preictal patterns of features. In this explorative study, we evaluated out-of-sample seizure prediction performance, and compared each combination of feature type and classifier. RESULTS: Among the evaluated methods, convolutional networks combined with wavelet coherence successfully predicted all out-of-sample seizures, without false alarms, on 15 patients, yielding 71% sensitivity and 0 false positives. CONCLUSIONS: Our best machine learning technique applied to spatio-temporal patterns of EEG synchronization outperformed previous seizure prediction methods on the Freiburg dataset. SIGNIFICANCE: By learning spatio-temporal dynamics of EEG synchronization, pattern recognition could capture patient-specific seizure precursors. Further investigation on additional datasets should include the seizure prediction horizon.
OBJECTIVE: Research in seizure prediction from intracranial EEG has highlighted the usefulness of bivariate measures of brainwave synchronization. Spatio-temporal bivariate features are very high-dimensional and cannot be analyzed with conventional statistical methods. Hence, we propose state-of-the-art machine learning methods that handle high-dimensional inputs. METHODS: We computed bivariate features of EEG synchronization (cross-correlation, nonlinear interdependence, dynamical entrainment or wavelet synchrony) on the 21-patient Freiburg dataset. Features from all channel pairs and frequencies were aggregated over consecutive time points, to form patterns. Patient-specific machine learning-based classifiers (support vector machines, logistic regression or convolutional neural networks) were trained to discriminate interictal from preictal patterns of features. In this explorative study, we evaluated out-of-sample seizure prediction performance, and compared each combination of feature type and classifier. RESULTS: Among the evaluated methods, convolutional networks combined with wavelet coherence successfully predicted all out-of-sample seizures, without false alarms, on 15 patients, yielding 71% sensitivity and 0 false positives. CONCLUSIONS: Our best machine learning technique applied to spatio-temporal patterns of EEG synchronization outperformed previous seizure prediction methods on the Freiburg dataset. SIGNIFICANCE: By learning spatio-temporal dynamics of EEG synchronization, pattern recognition could capture patient-specific seizure precursors. Further investigation on additional datasets should include the seizure prediction horizon.