Identifying rhythms of subthalamic neural oscillations in time-frequency domain.
Wang S., Green A., Stein J., Aziz T., Liu X.
We aimed to identify neural oscillations in the time-frequency representation of local field potentials recorded from the subthalamic nucleus. The time-frequency representation was normalised over the global mean and standard deviation global normalisation, or against the baseline period at each frequency, local normalisation. The cross-correlation between beta and gamma oscillations was enhanced by global normalisation. Furthermore, voluntary movement related amplitude changes in the gamma band and frequency modulation in the beta band were revealed by local normalisation. Thus global or local normalisation of time-frequency representation provides a reliable and effective way to identify oscillatory rhythms in subthalamic neural activity by reducing noise and increasing frequency discrimination. It can be used to enhance the detection of obscure or hidden neural oscillations and improve the sensitivity of post-hoc analysis.