Mapping cortical excitability in the human dorsolateral prefrontal cortex

Abstract

Background: Repetitive transcranial magnetic stimulation (rTMS) to the dorsolateral prefrontal cortex (dlPFC) is an effective treatment for depression, but the neural response to rTMS remains unclear. TMS with electroencephalography (TMS-EEG) can probe these neural effects, but variation in TMS-evoked potentials (TEPs) across the dlPFC are not well characterized and often obscured by muscle artifact. Mapping TEPs and artifacts across dlPFC targets is needed to identify high fidelity subregions that can be used for rTMS treatment monitoring. Objective: To characterize ′early TEPs′ anatomically and temporally close (20-50 ms) to the TMS pulse and associated muscle artifacts (<20 ms) across the dlPFC. We hypothesized that TMS location and angle would affect these early TEPs and that TEP size would be inversely related to muscle artifact. We sought to identify an optimal TMS target / angle for the group and asked if individualization would be beneficial. Methods: In 16 healthy participants, we applied single-pulse TMS to six targets within the dlPFC at two coil angles and measured EEG responses. Results: Early TEPs were sensitive to stimulation location, with posterior and medial targets yielding larger early TEPs. Regions with high early TEP amplitude had less muscle artifact, and vice versa. The best group-level target yielded 102% larger TEP responses compared to other standard targets. Optimal TMS target differed across subjects, suggesting that a personalized targeting approach could boost the early TEP by additional 36%. Conclusions: The early TEPs can be probed without significant muscle-related confounds in posterior-medial regions of the dlPFC. A personalized targeting approach may further enhance the signal quality of the early TEP.

Reliability of the TMS-evoked potential in dorsolateral prefrontal cortex

“…Relatedly, we hypothesized that larger EL-TEPs would be observed from stimulation of the posterior and medial dlPFC due to weaker muscle activation (and therefore smaller masking artifacts). Our findings confirmed these hypotheses, with posterior and medial targets eliciting larger EL-TEPs compared to anterolateral targets (4). However, stronger EL-TEP amplitudes at specific subregions importantly do not necessarily translate to higher reliability.…”

“…One such promising metric is derived from TMS coupled with electroencephalography (TMS-EEG), which produces a causal measure of cortical excitability (2,3). Here, the ‘ EL-TEP’ or the local and short-latency ( early, 20-60 ms) response to single TMS pulses applied to the dlPFC (4) represents an exciting interpretable neurophysiological metric with neural underpinnings (5). EL-TEPs may reflect aspects of prefrontal GABAergic inhibition and NMDA receptor function (6), but these studies need to be further validated and tested in the dlPFC.…”

“…Recent work in our lab investigated the effect of applying single TMS pulses at different dlPFC subregions on EL-TEP amplitude (4). We hypothesized that because the anterolateral portion of the dlPFC activates scalp muscle that induces large early artifacts in EEG (0-20 ms after a single pulse of TMS (14)) which are difficult to remove, the ‘true’ underlying neural response (e.g.…”

Neural effects of TMS trains on the human prefrontal cortex

“…We chose these time windows following Gogulski et al . (2023) 42 and because they are far enough after the interpolated time window (ending at 12 ms) and before time windows that are reported to include strong off-target sensory effects 16,21,22 . The early 20-50 ms time window captures our primary hypothesized component of interest, the early local TEP.…”

“…Because apparent amplitude of an averaged EEG waveform is not independent of latency variance, and the early TEP peaks are not yet well defined, we used a metric (area under the curve, AUC) to capture the full morphology of the LMFA waveform by aggregating the LMFA waveform over a time window rather than focusing on an instantaneous amplitude 40,41 . We chose these time windows following Gogulski et al (2023) 42 and because they are far enough after the interpolated time window (ending at 12 ms) and before time windows that are reported to include strong off-target sensory effects 16,21,22 . The early 20-50 ms time window captures our primary hypothesized component of interest, the early local TEP.…”