Pre-stimulus theta power is correlated with variation of motor evoked potential latency: a single-pulse TMS study

Abstract

There has been a growing interest in the role of pre-stimulus oscillations on cortical excitability in visual and motor systems. Prior studies focused on the relationship between pre-stimulus neuronal activity and TMS-evoked motor evoked potentials (MEPs) have reported heterogeneous results. We aimed to assess the role of pre-stimulus neural activity on the latency of MEPs, which might enhance our understanding of the variability of MEP signals, and potentially provide information on the role played by cortical activity fluctuations in the excitability of corticospinal pathways. Near-threshold single-pulse TMS (spTMS) was applied at random intervals over the primary motor cortex of 14 healthy participants while they sat passively, to trigger hand muscle contractions. Multichannel EEG was recorded during spTMS blocks. Spearman correlations between both the variation in MEP onset latencies and peak-to-peak MEP amplitudes, and the pre-stimulus power of EEG oscillations were calculated across participants. We found that the variation in MEP latency was positively correlated with pre-stimulus power in the theta range (4-7 Hz) in a broad time window (- 3.1 to - 1.9 s) preceding the spTMS generating the MEP. No correlation between pre-stimulus power in any frequency band and MEP amplitude was found. Our results show that pre-stimulus theta oscillations are correlated with the variation in MEP latency, an outcome measure determined by fiber conduction velocity and synaptic delays along the corticospinal tract. This finding could prove useful for clinicians using MEP latency-based information in pre- or intra-operative diagnostics of corticospinal impairment.

Dynamic relationships between spontaneous and evoked electrophysiological activity

“…Interestingly, the consistency of our spectral power results between different modalities and paradigms differs from previous findings, which showed task-specific relationships between pre-stimulus spectral power and poststimulus ERP components 29 , 79 , 81 . This may, in part, be due to the fact that we investigated the same neurophysiological variables during the transition from prestimulus to post-stimulus period; alternatively, they may be due to the use of time-domain signals to operationalize stimulus response, as we likewise observed task-specific spontaneous-evoked correlations in our ERP data.…”

The Common Rhythm of Action and Perception

“…By probing directly corticospinal excitability, Transcranial Magnetic Stimulation (TMS) studies provide more compelling evidence that neuronal oscillations actually entail cyclic modulations of the motor system excitability. Indeed, motor evoked potentials (MEPs) are modulated not only by the amplitude (İşcan et al, 2018;Keil et al, 2014;Mäki and Ilmoniemi, 2010;Schulz et al, 2014) but also by the pre-TMS phase of ongoing oscillations at both peripheral (i.e., muscle activity; Keil et al, 2014, van Elswijk et al, 2010 and cortical level (Berger et al, 2014;Keil et al, 2014;Khademi et al, 2018). The use of neuromodulation techniques, such as transcranial Alternating Current Stimulation (tACS), in combination with TMS, have further shown that entrained, in addition to ongoing, beta-band oscillations affect MEPs size in a phase-dependent manner (Guerra et al, 2016;Nakazono et al, 2016;Raco et al, 2016;Schilberg et al, 2018).…”

Differences in Power Spectral Densities and Phase Quantities Due to Processing of EEG Signals

“…In these applications, an array of signal processing methods have been applied to EEG signals over the past two decades many of which were summarized by Bashashati et al [13]. Using the processing methods, an array of features is usually extracted from the EEG signals [14][15][16][17][18][19][20]. The features are then used for various tasks such as statistical analysis or machine learning based methods [14][15][16][17][18][19][20][21].…”

“…Using the processing methods, an array of features is usually extracted from the EEG signals [14][15][16][17][18][19][20]. The features are then used for various tasks such as statistical analysis or machine learning based methods [14][15][16][17][18][19][20][21]. The EEG signal processing methods, in themselves, have a number of selection parameters, which have been found to be selected inconsistently across different studies.…”

“…In other signal processing domains, such as studies dealing with analyzing the effect of pre-stimulus EEG on motor evoked potentials (MEP) caused by Transcranial Magnetic Stimulation, methodological differences for PSD estimation, filtering, etc., have been found. For example,İşcan et al used Fast Fourier Transformation (FFT), which is non-parametric in nature, to estimate frequency band PSD in the pre-stimulus time-window starting from −4.5 s and −0.5 s (0 s indicates the time of stimulation) [19]. On the contrary, for a similar goal, Hussain et al used a parametric autoregressive model approach (Burg's method) on the time-window of 150 ms [18,25,30].…”