In recent years, new recording technologies have advanced such that oscillations of neuronal networks can be identified from simultaneous, multisite recordings at high temporal and spatial resolutions. However, because of the deluge of multichannel data generated by these experiments, achieving the full potential of parallel neuronal recordings also depends on the development of new mathematical methods capable of extracting meaningful information related to time, frequency and space. In this review, we aim to bridge this gap by focusing on the new analysis tools developed for the automated detection of high-frequency oscillations (HFOs, >40Hz) in local field potentials. For this, we provide a revision of different aspects associated with physiological and pathological HFOs as well as the several stages involved in their automatic detection including preprocessing, selection, rejection and analysis through time-frequency processes. Beyond basic research, the automatic detection of HFOs would greatly assist diagnosis of epilepsy disorders based on the recognition of these typical pathological patterns in the electroencephalogram (EEG). Also, we emphasize how these HFO detection methods can be applied and the properties that might be inferred from neuronal signals, indicating potential future directions.
Keywords: EEG; Epilepsy; Fast ripple; Gamma; HFO; High frequency oscillations; Macroelectrodes; Microelectrodes; Ripple; iEEG.
Copyright © 2017. Published by Elsevier Ltd.