On the time resolution of event-related desynchronization: a simulation study

Abstract

Objectives: To investigate the time resolution of different methods for the computation of event-related desynchronization/synchronization (ERD/ERS), including one based on Hilbert transform.Methods: In order to better understand the time resolution of ERD/ERS, which is a function of factors such as the exact computation method, the frequency under study, the number of trials, and the sampling frequency, we simulated sudden changes in oscillation amplitude as well as very short and closely spaced events.Results: Hilbert-based ERD yields very similar results to ERD integrated over predefined time intervals (block ERD), if the block length is half the period length of the studied frequency. ERD predicts the onset of a change in oscillation amplitude with an error margin of only 10-30 ms. On the other hand, the time the ERD response needs to climb to its full height after a sudden change in oscillation amplitude is quite long, i.e. between 200 and 500 ms. With respect to sensitivity to short oscillatory events, the ratio between sampling frequency and electroencephalographic frequency band plays a major role.Conclusions: (1) The optimal time interval for the computation of block ERD is half a period of the frequency under investigation. (2) Due to the slow impulse response, amplitude effects in the ERD may in reality be caused by duration differences. (3) Although ERD based on the Hilbert transform does not yield any significant advantages over classical ERD in terms of time resolution, it has some important practical advantages. q

Evidence for Enhanced Multisensory Facilitation with Stimulus Relevance: An Electrophysiological Investigation

“…Thus, both ERPs and induced oscillations were modulated by stimulus relevance at a late stage of neural processing, which coincided with preparation and initiation of motor action. However, given the low temporal resolution of time-frequency transforms, particularly for low frequency oscillations in the alpha and beta range [69], our results raise a further fundamental question of whether late oscillations, after 200 ms post stimulus, are driving multisensory facilitation or whether they are a consequence of earlier integrative processes. Our results suggest that the neural synchronization driving behavioral multisensory facilitation may not only involve early processes, within the first 170 ms in sensory specific regions, but also the late activation of neural networks and assemblies engaged in stimulus selection, decision making, and the preparation and initiation of motor actions.…”

Event-related alpha and theta responses in a visuo-spatial working memory task

“…And third, the present effect is sustained, and even increases in strength, throughout the entire delay period, lasts 4 s. We feel that it is more likely that the early, i.e. pre-target onset of this effect results from the poor temporal resolution of the IBP analysis method (see Knösche and Bastiaansen, 2002 for an elaborate discussion of the temporal resolution of IBP analysis).…”

Theta Responses Are Involved in Lexical—Semantic Retrieval during Language Processing

“…The temporal resolution of timefrequency analysis is inherently relatively poor compared to ERP measures (e.g., Knosche & Bastiaansen, 2002), with lower frequencies having poorer temporal resolution (in the order of a few hundred milliseconds) than higher frequencies. In addition, the use of sliding windowing techniques as in the present wavelet analysis leads to some additional temporal smearing of effects.…”