Does the morphology of high-frequency (100–500 Hz) brain oscillations change during epileptic seizures?

Transient high-frequency (100-500 Hz) oscillations (HFOs) recorded directly from the surface of the human brain are emerging as a potential biomarker for epileptogenic brain tissue. Whether the morphology of these events can be used to understand the process of seizure generation is unknown. In this experiment, we used supervised learning techniques in an attempt to distinguish HFOs occurring during versus outside of seizures in five patients implanted with intracranial micro-and macroelectrodes as part of routine evaluation for epilepsy surgery. We trained three classifiers using logistic regression, k-nearest neighbors, and support vector machines, respectively, and assessed their performance using the F1 measure in conjunction with permutation testing. All of the classifiers produced a low number of true positives relative to false positives and false negatives, but two of the classifiers performed slightly better than chance in certain patients. These results suggest that ictal HFOs are difficult to distinguish from those occurring interictally, and that a marked change in HFO morphology is not likely to be associated with seizure generation.