Fig 1.
Experimental design and procedures.
(A) Timeline of sessions and procedures for a single experiment (note that durations provided refer to the time spent on each particular experimental stage and not just the mere measurement time). (B) Position of surface electrode pairs for finger-specific PES on the digit I, III, and V of the right hand. (C) Schematic depiction of phase-targeted PES, resulting in the arrival of somatosensory input in S1 at the peak, trough, or random phases of the sensorimotor mu-alpha rhythm. PES was applied at an interval prior to the respective phase that reflected the individual N20 latency of the SEP.
Fig 2.
Somatosensory evoked potentials (SEP).
(A) The grand average SEP waveform across all subjects (n = 16), extracted from the N20 and P25 clusters, respectively. (B) Grand average topographies calculated from the individual time-locked averages time series +/- 2 ms with respect to each individual subject’s N20 and P25 peak, respectively. (C) Comparisons of N20 and P25 voltage values pre- and post-intervention, where the left-hand side of the violins represent the pre-intervention and the right-hand side the post-intervention for a given condition. (D) Comparisons of the N20 and P25 amplitudes before (SEPpre) vs. after (SEPpost) the EEG-triggered PES intervention expressed in percent change of SEPpost from SEPpre. For both C and D, the outer shape of the violin represents the kernel density estimation, and gray dots within the violins represent individual subject averages per condition. Within each violin plot a box plot is provided (marking with its upper and lower border the 25% and 75% percentiles), with white circles representing the condition median, horizontal gray lines representing the condition mean, and vertical bars extending from there an additional 1.5 times the interquartile range.
Table 1.
Mean ± SD for raw (μV) and normalized (% change of post from pre PES-intervention SEPs) SEP N20 and P25 amplitudes for all phase targets (peak, trough, random).
Fig 3.
Sensory Threshold (ST) violin plots.
(A) Split violin graphs where the left-hand side of the violins represent STpre and the right-hand side STpost for a given Phase condition. (B) comparisons of the values after the EEG phase-triggered PES intervention between Phase conditions (STpost) expressed in percent change from those before (STpre.) For both A and B, the outer shape of the violin represents the kernel density estimation, and gray dots within the violins represent individual subject averages per condition. Within each violin plot a box plot is provided (marking with its upper and lower border the 25% and 75% percentiles), with white circles representing the condition median, horizontal gray lines representing the condition mean, and vertical bars extending from there an additional 1.5 times the interquartile range.
Table 2.
Mean ± SD for raw (mA) and normalized (% change of post from pre PES-intervention STs) sensory thresholds (ST) for all phase targets (peak, trough, random).
Fig 4.
Targeting validation for EEG phase-triggered PES intervention.
Colors of elements indicate the Phase target conditions peak (red), trough (blue), and random (green). (A) Grand average time series across all subjects for the three Phase conditions, individually time-locked to the respectively forecasted phase target (bold vertical line at 0 ms), marking the estimated arrival of afferent input in S1. The average position of the PES triplet, which precedes the phase target by the individual N20 latency and thus varies across subjects, is indicated by the three thinner vertical lines. (B) Average phase angles targeted per Phase condition (as estimated from the preceding mu-alpha cycle to avoid corruption by PES artifacts and SEPs, with colored circles on the unitary cycle representing the individual phase averages per target condition and vectors representing the condition-wise grand average phase angles (by their direction) and phase consistencies (by their magnitude) across subjects. (C) Grand average power spectra across all subjects for the -500 to -50 ms pre-target (and pre-PES) window for the three Phase conditions.