High-Voltage Spindles detection from EEG signals using recursive synchrosqueezing transform (original) (raw)

High-Voltage Spindles detection from EEG signals using recursive synchrosqueezing transform

This paper proposes a new detection technique applied on electroencephalogram (EEG) signals which were measured on Parkinsonian rats during an experiment. Our technique uses a recursive filter-bank-based implementation of the short-time Fourier transform that is sharpened using several time-frequency reassignment methods. A detector is then applied to the obtained representation to allow an identification of the HVS signals which are specific to subjects with neurodegeneratives diseases. Our results show an improvement of the state of the art while paving the way of a real-time implementation.