Claims 45
From the paper Parametrizing alternating current stimulation for neuromodulation · claims file · checked 2026-10-21: ready
- efield-1min-medians
The electric fields estimated from the stimulation artifact in the 1-minute experiments had medians of 0.77, 1.42, 2.86 and 6.04 mV/mm for 50, 100, 200 and 400 uA.
The magnitude of electric fields was assessed from sACS artifacts in electrophysiological recordings, and had medians of 0.77 ± 0.29, 1.42 ± 0.60, 2.86 ± 1.20, and 6.04 ± 1.68 mV/mm for 50, 100, 200, 400 μA sACS amplitudes, respectively
Results 3.1, paragraph 1; Figure 1CHow it will be judged
Method from Methods 2.8. For each rat and amplitude, every channel of every trial of that amplitude is band-pass filtered at the stimulation frequency (5, 10, 20 or 40 Hz); the peak voltage of each stimulation cycle is averaged over the stimulation period per channel; differences between neighbouring electrodes along each array axis are divided by the 0.5 mm spacing (Methods 2.3), the two orthogonal fields combined by vector addition, outliers removed by Grubbs' test (alpha 0.05), and the field averaged over the frequency trials of the amplitude. Tolerances are the paper's stated spread. Open, with readings run (a) the median over rats of each rat's average field, (b) the median over all channel-pair estimates of all rats (Fig. 1C left shows these); "maximal" field per axis read as (i) each neighbour pair or (ii) the largest pair per axis and rat. The band-pass filter is not named; a zero-phase Butterworth band of +/- 1 Hz around the stimulation frequency on the LFP stream is the standard choice and is stated as one. Reproduced when all four medians are within tolerance under one reading and rise with amplitude.
- efield-20min-median
The median electric field in the 20-minute 100 uA experiments was 1.65 mV/mm.
The median electric field produced for these experiments was 1.65 ± 0.21 mV/mm
Results 3.4, paragraph 1; Figure 1DHow it will be judged
As efield-1min-medians (Methods 2.8, 40 Hz band), over the 13 rats, with the same readings; tolerance the stated spread.
- fem-depth-falloff
In the finite element model the electric field falls to about half of its value at the recording electrodes about 70 um deeper.
Within the model, the electric field dropped to ~50% of the intensity measured at the recording electrodes at a further ~70 µm in depth
Results 3.1, paragraph 2; Figure 2C legendHow it will be judged
Sim4Life Electro Ohmic Quasi-Static solver on the ViZOO Big Male Rat model with resected tissue removed, 1.5 x 1.5 x 0.1 mm gold electrodes, grid 0.5 mm (0.1 mm between electrodes), scaled so that the RMS field in the 2 x 2 mm array space 300 um deep matches the measured fields (Methods 2.8). The dataset supplies only the measured fields.
- fem-lognormal-fits
Cerebral electric field RMS values in the model follow log-normal distributions with shape 1.28 at every amplitude and scale 0.217, 0.400, 0.807 and 1.70 for 50, 100, 200 and 400 uA.
The cerebral electric field root mean square (RMS) values were fitted with log-normal distributions, with shape, location, and scale parameters of 1.28, 8.54E-03, 2.17E-01 for 50 µA; 1.28, 1.57E-02, 4.00E-01 for 100 µA; 1.28, 3.17E-02, 8.07E-01 for 200 µA; and 1.28, 6.70E-02, 1.70 for 400 µA
Results 3.1, paragraph 2; Figure 2DHow it will be judged
scipy.stats lognorm fitted to the cerebral RMS values of the model of fem-depth-falloff (Methods 2.8); units as the paper's (mV/mm, read from the 1 mV/mm threshold). The locations (8.54E-03 to 6.70E-02) are reported alongside.
- fem-ellipsoid-axes
Ellipsoids fitted to the hemispheric volumes with fields of at least 1 mV/mm have axes of 7.37, 2.46, 1.58 mm (50 uA), 7.37, 3.32, 2.22 mm (100 uA), 7.86, 4.73, 3.28 mm (200 uA) and 8.80, 6.57, 4.35 mm (400 uA).
Ellipsoids were fit to thresholded hemispheric volumes experiencing electric field magnitude of at least 1 mV/mm, with primary, secondary, and tertiary axes lengths of 7.37, 2.46, and 1.58 mm for 50-µA ACS; 7.37, 3.32, and 2.22 for 100-µA ACS; 7.86, 4.73, and 3.28 mm for 200-µA ACS; and 8.80, 6.57, and 4.35 mm for 400-µA ACS
Results 3.1, paragraph 2; Figure 2EHow it will be judged
RMS field of the model thresholded at 1 mV/mm; pyvista.principal_axes on the surface points, standard deviations along each axis times 1.96 as radii (Methods 2.8). Each of the twelve axis lengths within 10% of the paper's value; "lengths" read as (a) the radii, (b) twice the radii, both run.
- units-1min
In the 1-minute experiments a median of 5 neurons was detected per channel and 66 per rat, of which 71% (405 of 586 cumulatively) fired at 0.25 Hz or more in at least half of the 10-second windows and were analysed.
Across rats, a median of 5.00 ± 0.17 neurons were detected per recording electrode channel using SpyKING CIRCUS ... This amounted to 66.00 ± 19.60 neurons per rat, of which 71.43 ± 10.39% (or, cumulatively, 405/586) had FRs ≥0.25 Hz in at least half of 10-second windows (of baseline and stimulation periods) and were, therefore, included in subsequent analysis
Results 3.1, paragraph 3; Figure 3A, BHow it will be judged
Sorting and inclusion as in the opening comment, under readings S1 and S2. The per-channel tolerance is 10% (the stated 0.17 is narrower than the integer steps of a count); the others are the stated spread, or 10% for the totals. A neuron counts as included when it passes in (a) at least one trial, (b) at least half of its trials; both run. Spike sorting has settings the paper does not give, so counts can differ for reasons other than the data; the result is reported for each reading.
- units-20min
In the 20-minute experiments a median of 44 neurons was detected per rat and 4 per channel; 52% (343 of 620) passed the 0.25 Hz criterion and were analysed.
A median of 44.00 ± 10.36 neurons were detected per rat, with 4.00 ± 0.60 neurons per electrode channel. Of these neurons, 51.61 ± 10.62% (or, cumulatively, 343 of 620) had FRs ≥0.25 Hz in at least half of all 1-minute windows across baseline and stimulation periods
Results 3.4, paragraph 1How it will be judged
As units-1min. The Results say 1-minute windows here while Methods 2.6 and phaseextract.py use 10-second windows; readings (a) 10-s windows over the 20 min baseline and the 20 min stimulation with more than 2 spikes in each (code), (b) 1-min windows, both run, each under S1 and S2.
- baseline-fr
Median baseline firing rates were 1.20 Hz in the 1-minute and 1.05 Hz in the 20-minute experiments.
Neurons from 1-min experiments had a median firing rate of 1.20 ± 2.11 Hz, and those from 20-min experiments had a median firing rate of 1.05 ± 4.92 Hz
Supplementary Figure S1.1; Results 3.1, paragraph 4 and 3.4, paragraph 1How it will be judged
Mean baseline FR of each included neuron (Figure S1.1, all neurons of all rats), median over neurons. 10% tolerance, as the stated spreads (2.11 and 4.92 Hz) are wider than the values and do not bound a median. For the 1-min experiments a neuron enters once per included trial (a) or once with its mean over trials (b); both run.
- baseline-ppc
Median baseline entrainment (sqrt|PPC|) was 0.12, 0.12, 0.11 and 0.11 to the 5, 10, 20 and 40 Hz waveforms in the 1-minute experiments and 0.14 to 40 Hz in the 20-minute experiments.
Neurons from 1-min experiments had median √|PPC| of 0.12 ± 0.05, 0.12 ± 0.07, 0.11 ± 0.10, 0.11 ± 0.09 to 5, 10, 20, and 40 Hz endogenous oscillatory frequencies, respectively, and those from 20-min experiments had a median √|PPC| of 0.14 ± 0.06 to 40 Hz frequencies
Supplementary Figure S1.2; Results 3.1, paragraph 4How it will be judged
Baseline spike phases taken from the stimulating waveform at the same offset (phaseextract.py), sqrt|PPC| per included neuron and trial, median over neurons by stimulation frequency; tolerances the stated spread. For the 1-min experiments a neuron enters once per included trial (a) or once with its mean over the trials of a frequency (b); both run.
- table1-any-change
Across 1-minute trials about 75% of included neurons per rat were significantly affected (FR or entrainment change), and some neurons were affected under every frequency and amplitude combination.
Across trials, about 75% of neurons were affected (amplitude/frequency-specific percentages are listed in Table 1 )
Results 3.1, paragraph 3; Table 1How it will be judged
Table 1 gives, per combination, the median over rats and 95% CI of the percentage of included neurons with a significant change (legend); the overall value is its last cell. Its tolerance is the printed CI half-width. "Overall" read as (a) the median of all rat-by-combination percentages, (b) the median of the 16 combination medians; both run, each under W1/W2 and P1/P2. Reproduced when the overall value is within tolerance and the median is above zero in all 16 combinations.
- fr-change-more-common-than-timing
In 1-minute trials a change in FR was far more common than a change in spike timing (about 70% against under 10% of neurons).
A change in FR was more frequent than a change in spike timing (approximately 70% vs. <10%; Table 1 )
Results 3.1, paragraph 3; Table 1How it will be judged
Overall cells of Table 1, readings as table1-any-change. Reproduced when the four values are within tolerance (the CI half-widths) and the share with any FR change is larger than the share with any sqrt|PPC| change, the latter below 10%.
- fr-change-mostly-negative
The FR change in 1-minute trials was mostly a decrease, about 60% of neurons inhibited against under 10% excited.
The majority of neurons responding to sACS decreased their FRs during stimulation (60% of neurons were inhibited by sACS while <10% were excited, Table 1 )
Results 3.3, paragraph 1; Table 1; Figures 4C, D and 10How it will be judged
The text gives 60% where Table 1's overall cell is 69.23% (95% CI half-width 12.64, so 56.59 to 81.87%). Reproduced when the overall median share with a significant FR decrease lies in 56.59 to 81.87% and the share with an FR increase is below 10%, under the readings of table1-any-change.
- ppc-change-mostly-increase
Of the neurons whose spike timing changed in 1-minute trials, nearly all increased their entrainment (about 4%), and very few decreased it (under 0.5%).
whereas the vast majority of spike timing affected neurons showed increases in entrainment (4%, Table 1 ... and only a few neurons exhibited negative Δ√|PPC| values (<0.5%; Table 1
Results 3.1, paragraph 4; Table 1; Figures 6 and 7How it will be judged
Overall cells of Table 1 (4.17 +/- 2.61% increase, 0.00 +/- 0.44% decrease), readings as table1-any-change. Reproduced when the share with a sqrt|PPC| increase is within 4.17 +/- 2.61%, the share with a decrease is below 0.5%, and the increase is larger.
- table2-magnitudes
Among affected neurons in 1-minute trials the median changes were +0.32 and -0.13 in sqrt|PPC| and +0.85 and -0.91 Hz in FR.
How it will be judged
Per Table 2's legend, the median and 95% CI of the change from baseline to stimulation among neurons significant in that category; tolerance the CI half-width. Readings for the pooling (a) all affected neuron-trials pooled, (b) per-rat medians then the median; each under W, P and G readings (G applies as the values may be cleaned before tabulation).
- fr-change-within-2hz
In 1-minute trials FR changes stayed within about 2 Hz of baseline, while entrainment changes were pronounced, Delta sqrt|PPC| reaching 0.55.
FR remained within ~2 Hz of baseline level ( Table 2 ), 1.20 ± 2.11 Hz ( Supplementary Fig. S1.1 ), whereas entrainment levels achieved were pronounced (Δ√|PPC| of up to 0.55
Results 3.1, paragraph 4; Table 2How it will be judged
Table 2 medians as in table2-magnitudes, per combination. Reproduced when every combination's median FR change (excited and suppressed) has absolute value below 2 Hz and the largest median sqrt|PPC| increase is 0.55 +/- 0.07 (Table 2's CI) at 40 Hz, 400 uA.
- fig4-distributions
With increasing amplitude and frequency, the distribution of neurons' changes broadens, shifting toward stronger entrainment and slight suppression of firing.
With increasing sACS amplitude (A) and frequency (B), neurons exhibit a broader distribution of responses, with a shift toward stronger entrainment from baseline and a slight suppression of their activity (C, D)
Figure 4 legendHow it will be judged
All included neuron-trials (legend, "all neurons"), Delta sqrt|PPC| and Delta FR grouped by amplitude (A, C) and by frequency (B, D). The paper fits density functions and names no measure of breadth or shift; readings for breadth (a) standard deviation, (b) interquartile range. Reproduced when, under one reading, the breadth of Delta sqrt|PPC| is larger at 400 than at 50 uA and at 40 than at 5 Hz, the median Delta sqrt|PPC| is higher at 400 than 50 uA and at 40 than 5 Hz, and the median Delta FR is below zero at 400 uA and at 40 Hz.
- lmer-ppc-increase-interaction
Among neurons with increased entrainment, the interaction of amplitude and frequency positively affected the degree of entrainment (LMER), with no significant main effects.
An LMER revealed that the interaction between amplitude and frequency positively affected the degree of entrainment ( Table 3 , p = 0.02
Results 3.2, paragraph 1; Table 3How it will be judged
Methods 2.6. Delta sqrt|PPC| of neurons with significant increases, on amplitude (uA) and frequency (Hz) as numbers with their interaction (Table 3 gives one coefficient each), random intercepts for neuron within subject, lme4 with Satterthwaite df (Table 3 prints df), Grubbs-cleaned (G readings). Reproduced when the interaction is positive with p < 0.05 and the amplitude and frequency main effects have p >= 0.05 (Table 3, 0.683 and 0.410), under one reading of W, P and G; partially when the interaction is positive and significant but a main effect is too.
- lmer-ppc-decrease-amplitude
Among neurons with decreased entrainment, only amplitude significantly affected the decrease, larger amplitudes giving smaller decreases.
Only amplitude was found to significantly affect decreases in entrainment ( Table 3 , p = 0.01; Fig. 7 ; Supplementary Tables S1.2–S1.5 ), with increasing amplitude yielding smaller magnitudes of entrainment decrease
Results 3.2, paragraph 1; Table 3How it will be judged
Model as lmer-ppc-increase-interaction, on the Delta sqrt|PPC| (negative) of neurons with significant decreases. Reproduced when the amplitude effect is positive with p < 0.05 and frequency and interaction have p >= 0.05, under one reading.
- lmer-fr-increase-no-dependence
The FR change of neurons that increased their FR did not depend on stimulation parameters.
Neurons that increased their FR during sACS did not show a dependence of their induced FR change on the sACS parameters ( Table 3 )
Results 3.3, paragraph 1; Table 3How it will be judged
A claim of absence, judged on the paper's test, the LMER of Methods 2.6 (as lmer-ppc-increase-interaction) on Delta FR of neurons with significant FR increases, at p <= 0.05. Reproduced when amplitude, frequency and their interaction all have p > 0.05 (Table 3, 0.434, 0.408, 0.463); not reproduced when any is significant. Estimates are reported either way, under W1 and G1 (the readings that do not touch the test as written are shown as variants).
- lmer-fr-decrease
For suppressed neurons, amplitude and frequency had independent and interactive effects on the decrease in FR (larger amplitude and frequency, deeper suppression).
an LMER revealed that amplitude and frequency had independent and interactive effects on the decrease in FR ( Table 3 , independent effect of amplitude
Results 3.3, paragraph 1; Table 3 and legendHow it will be judged
Model as lmer-ppc-increase-interaction on Delta FR of neurons with significant FR decreases. The Results print p = 0.001 for the interaction where Table 3 gives 1.31E-02 and its legend p = 0.01; the table is used. Reproduced when amplitude and frequency are negative and the interaction positive, each with p < 0.05, under one reading.
- fig6a-ppc-frequency-400ua
At 400 uA the increase in entrainment grew with stimulation frequency, 5.10e-3 Delta sqrt|PPC| per Hz.
The 400 µA linear regressions showed significant increases in Δ√|PPC| with increasing frequency at 5.10 × 10 −3 Δ√|PPC| per Hz, p = 0.002)
Figure 6A; Results 3.2, paragraph 2; Table S1.2How it will be judged
Least-squares regression (scipy.stats.linregress) of Delta sqrt|PPC| of neurons with significant increases at 400 uA on frequency (5, 10, 20, 40 Hz), readings R1/R2, W, P and G. Reproduced when the slope is positive with p <= 0.05 and within tolerance under one reading; partially when positive and significant but outside tolerance. Of the other amplitudes Table S1.2 reports none significant. The Results cite Table S1.4 for this regression, but the slope and p are in Table S1.2.
- fig6b-ppc-amplitude-5-40hz
At 5 and 40 Hz the increase in entrainment grew with amplitude, by 7.22e-4 and 7.63e-4 Delta sqrt|PPC| per uA.
with increasing sACS amplitudes, the degree of entrainment increased across 5 and 40 Hz stimulation frequencies (p = 0.004 and 0.0002, respectively); growing at a rate of 7.22 × 10 −4 and 7.63 × 10 −4 Δ√|PPC| per μA
Figure 6B; Results 3.2, paragraph 2; Table S1.3How it will be judged
As fig6a-ppc-frequency-400ua, regressing on amplitude (50 to 400 uA) at each frequency. Reproduced when both slopes are positive with p <= 0.05 and within tolerance under one reading; partially when one holds.
- fig7-ppc-decrease-regressions
For neurons with decreased entrainment, only the 400 uA regression across frequency (-1.74e-3 per Hz) and the 5 Hz regression across amplitude (2.67e-4 per uA) were significant.
Only the 400-µA regression (A) across frequencies and the 5-Hz regression across amplitudes (B) showed significance, with -1.74 × 10 −3 Δ√|PPC| per Hz and 2.67 × 10 −4 Δ√|PPC| per µA (p = 0.04 and 0.005, respectively)
Figure 7A, B legend; Table S1.2, S1.3How it will be judged
As fig6a-ppc-frequency-400ua, on the Delta sqrt|PPC| of neurons with significant decreases. Reproduced when both named slopes have the paper's sign, p <= 0.05 and are within tolerance under one reading, and no other of the eight regressions is significant; partially when the two named slopes hold but another regression is also significant.
- fig8a-excited-fr-frequency-100ua
At 100 uA the FR increase of excited neurons grew with frequency, 3.65e-2 Hz per Hz.
Linear regressions of 100 μA sACS response data versus sACS frequency showed increases in the magnitude of FR increase in excited neurons with increasing frequency (3.65 × 10 −2 ΔHz per Hz, p = 0.0003)
Figure 8A; Results 3.3, paragraph 1; Table S1.2How it will be judged
As fig6a-ppc-frequency-400ua, on Delta FR of neurons with significant FR increases at 100 uA.
- fig8b-excited-fr-amplitude-5hz
At 5 Hz the FR increase of excited neurons grew with amplitude, 2.22e-3 Hz per uA.
The 5 Hz response regression across amplitudes ( Fig. 8B ; Supplementary Table S1.3 ) showed an increase in the magnitude of FR increase of excited neurons with increasing amplitude (2.22 × 10 −3 ΔHz per μA, p = 0.002)
Figure 8B; Results 3.3, paragraph 1; Table S1.3How it will be judged
As fig6b-ppc-amplitude-5-40hz, on Delta FR of neurons with significant FR increases at 5 Hz.
- fig9a-suppressed-fr-frequency
For suppressed neurons at 50 and 100 uA, FR decreased further with frequency, by -8.33e-3 and -1.19e-2 Hz per Hz.
The 50 and 100 µA regressions across frequencies (A) showed greater degrees of suppression in FRs with increasing frequency (-8.33 × 10 −3 and -1.19 × 10 −2 ΔHz per Hz, p = 0.0004 and <0.0001, respectively)
Figure 9A legend; Results 3.3, paragraph 2; Table S1.2How it will be judged
As fig6a-ppc-frequency-400ua on Delta FR of neurons with significant FR decreases. The Results print "50 and 10 μA"; the legend's 100 uA is used (10 uA was not tested). Reproduced when both slopes are negative, p <= 0.05 and within tolerance under one reading; partially when one holds.
- fig9b-suppressed-fr-amplitude-5hz
Only at 5 Hz did suppressed neurons' FR decrease further with amplitude, -1.36e-3 Hz per uA.
Only the 5-Hz regression (B) showed increased suppression with increasing amplitude (-1.36 × 10 −3 ΔHz per µA, p = 0.0008)
Figure 9B legend; Results 3.3, paragraph 2; Table S1.3How it will be judged
As fig6b-ppc-amplitude-5-40hz. Reproduced when the 5 Hz slope is negative, p <= 0.05 and within tolerance, and the 10, 20 and 40 Hz regressions are not significant, under one reading; partially when the 5 Hz slope holds and another is also significant.
- fig10a-all-fr-frequency
Taking all FR-changed neurons together, suppression deepened with frequency at 50, 100 and 200 uA (-1.02e-2, -1.33e-2, -1.47e-2 Hz per Hz), while it was largely maintained across amplitudes.
When considering grouped FR changes, most amplitudes of stimulation elicited further suppression with increasing stimulation frequency (at rates of -1.02 × 10 −2 , -1.33 × 10 −2 , and -1.47 × 10 −2 ΔHz per Hz for 50, 100, and 200 µA trials; p = 0.0003, <0.0001, and 0.0001, respectively
Figure 10 legend; Results 3.3, paragraph 2; Tables S1.2, S1.3How it will be judged
As fig6a-ppc-frequency-400ua on Delta FR of all neurons with a significant FR change (increase or decrease). Reproduced when the three slopes are negative, p <= 0.05 and within tolerance under one reading; partially when two hold. The amplitude regressions (Fig. 10B) are reported; "largely maintained" sets no test.
- lfp-1min-no-change
One minute of sACS produced no significant change in LFP power or LF-HF PAC from before to after stimulation.
LFP analysis did not show significant changes in power or LF-HF PAC before and after 1-min sACS ( Supplementary Tables S1.6–S1.7 )
Results 3.1, paragraph 4; Tables S1.6, S1.7How it will be judged
A claim of absence, judged on the paper's test, the ANOVA tables of LMERs (Tables S1.6, S1.7), one per response category (increase or decrease in FR or sqrt|PPC|), of log power (or order-normalized MI) on timepoint (baseline or post-sACS), band (or band pair), amplitude and frequency with all interactions, on channels proximal to affected neurons. The S1.7 legend names six band pairs (theta, alpha, beta to low and high gamma, no delta pairs), which fits its 5 df. The S1.6 legend names five bands (theta to high gamma) where its 5 df imply six; readings (a) the five named, (b) all six bands of Methods 2.7. Reproduced when the timepoint term and every interaction containing it have p >= 0.05 in all eight models; not reproduced when any is below. Methods 2.5 say LFP was analysed only around the 20-min trials, so the windows are open; read as the 1 min before onset and the 1 min after offset in 2-s (power) and 10-s (MI) windows, with the L and M readings. Effects reported either way.
- sham-no-fr-change
Under sham stimulation no neuron changed its FR in any of the 16 paradigms, and one or two neurons were spuriously detected as changing entrainment in some, without the trends seen under true sACS.
Under sham sACS, no neurons showed significant changes in firing rate across any of the 1-minute baseline to sham stimulation paradigms ( Supplementary Table S3.1 ). One or two neurons were spuriously detected as having significantly changed entrainment levels
Results 3.1, paragraph 5; Table S3.1How it will be judged
Methods 2.7, the 80-min sham recording cut into sixteen 3-min blocks, each assigned one amplitude and frequency, its first minute as baseline and its second as sham stimulation, through the 1-min pipeline; phases taken from a sinusoid at the assigned frequency, as no stimulating waveform exists (the phase offset does not change PPC). The phase reference for sham is not given; readings (i) the sinusoid, (ii) the recorded stimulation monitor channels of the sham recording, if the dataset holds them, as phaseextract.py uses for true sACS. Block assignment is random in the paper; it changes only the phase-reference frequency, and the 16 blocks are run in order (a standard choice). The FR part is a claim of absence judged on the paper's test, the FR permutation test at p <= 0.05; reproduced when no neuron in any block is significant, not reproduced otherwise, with counts reported. The sqrt|PPC| part holds when each block has at most two significant neurons in each of the two sqrt|PPC| categories (Table S3.1 shows 0 to 2 per category, and 3 in one block, 200 uA 5 Hz, counting both).
- sham-lfp-baseline
Without stimulation, LFP band power and modulation indices stayed stable over 80 minutes under propofol, for example delta power 1.71e-6 and delta to low-gamma MI 1.17e-3.
In 80-minute LFP recordings under propofol anesthesia without sACS perturbation, power was maintained at 1.71E-06 ± 6.94E-07 mV/Hz, 9.70E-07 ± 3.89E-07 mV/Hz, 6.66E-07 ± 2.72E-07 mV/Hz, 3.03E-07 ± 9.35E-08, 8.40E-08 ± 1.74E-08 mV/Hz, and 3.61E-08 ± 3.50E-09 mV/Hz for δ, θ, α, β, low-γ, and high-γ band activity
Results 3.7, paragraph 2 and 3.8, paragraph 2How it will be judged
Methods 2.7, the average and standard deviation of band power and of the raw MI in 1-min windows every 10 min of the sham recording; tolerances the stated SD. Open details, run as readings, power windows (a) 3 s, as the sham paragraph says, (b) 2 s, as for the 20-min analysis it claims to copy; units (the text prints mV/Hz) read as power spectral density in (i) mV^2/Hz, (ii) V^2/Hz; channels, all 16 averaged (none are singled out). The MI readings M1/M2 apply. Reproduced when, under one reading, all six power values and all eight MI values are within tolerance.
- 20min-during-percentages
Over the 20 minutes of sACS, 4% of neurons were entrained and 2% decreased their entrainment, while 54% increased and 34% decreased their FRs; on the whole neurons were excited.
4% of all neurons were entrained, while in 2% of all neurons √|PPC| decreased ( Table 4 ; Supplementary Fig. S2.1 ). When considering the FRs of neurons, 54% showed increased FRs, while 34% decreased their FRs
Results 3.5, paragraph 1; Table 4; Supplementary Figures S2.1, S2.2How it will be judged
The 20 min of stimulation against the 20 min of baseline (Methods 2.6), the pooled percentages of Table 4 over all included neurons of the 13 rats. Readings S, W (including the code's 55-s FR null) and P. Reproduced when, under one reading, all five are within tolerance (10%, as Table 4 gives no spread for pooled values) and more neurons increase than decrease FR. The per-rat medians of Table 4 (91.89, 0.00, 0.00, 63.41, 24.39%) are reported alongside.
- 20min-post-percentages
In the 10 minutes after the 20-minute sACS, 2% of neurons kept increased entrainment and 4% had decreased entrainment; 51% had decreased and 39% increased FRs, so excitation during sACS turned to suppression after it.
2% of all neurons maintained imposed spiking patterns (i.e., showed increases in entrainment from baseline after stimulus offset) ... Finally, half of all neurons (51%) had decreased firing post-stimulation ... while 39% showed increased FRs post-sACS ... excitation during sACS thus turned to suppression post-sACS
Results 3.6, paragraph 1; Table 4; Supplementary Figures S2.3, S2.4How it will be judged
The first 10 min after offset against the last 10 min before onset (Methods 2.6; nulldistribution.py), pooled over the 343 included neurons. Readings S and P. Reproduced when all five are within tolerance (10%) and more neurons decrease than increase FR after offset, while 20min-during-percentages shows the reverse during stimulation.
- 20min-magnitudes
Affected neurons changed sqrt|PPC| by +0.04 and -0.07 during 20-minute sACS and by +0.07 and -0.04 after it, with FR changes of +0.91 and -0.30 Hz during and +1.53 and -0.43 Hz after.
How it will be judged
Medians over the affected neurons of each category, pooled across rats (Table 4, with its 95% CI as tolerance); the same readings as the two percentage claims. The per-rat medians of Table 4 are reported alongside.
- power-entrained-channels
On the 12 channels near entrained neurons, log power fell from before to after 20-minute sACS in the delta, theta, alpha, beta and low-gamma bands.
For channels near entrained neurons, significant decreases in log power observed across δ, θ, α, β, and low-γ frequency bands (emmean = -0.12, -0.15, -0,13, -0.06, and -0.04
Results 3.7, paragraph 1; Figure 11B legend; Table S2.1How it will be judged
Methods 2.7 with readings L1/L2, B1/B2, N1/N2, F1/F2. Channels chosen from the neurons significant over the full 20 min of stimulation; LFP decimated to 1.2 kHz; FFT of non-overlapping 2-s windows over the 2 min before onset and after offset, mean power per band, log-transformed. Reproduced when, under one reading, the five bands decrease with p < 0.05 after correction and high gamma does not, and the contrasts are within tolerance (the SE printed in the Figure 11B legend for each response; for excited channels, whose legend prints 0.00, the 4.78E-03 of Table S2.1; channel counts 10%); partially when the signs and significance hold but the contrasts fall outside.
- power-reduced-entrainment-channels
On the 6 channels near neurons with decreased entrainment, log power rose in the alpha, beta, low-gamma and high-gamma bands.
whereas near decreases in √|PPC|, channels showed increases in the log power of α, β, low-γ, and high-γ frequency bands (emmean = 0.05, 0.09, 0.09, and 0.09
Results 3.7, paragraph 1; Figure 11B legend; Table S2.1How it will be judged
As power-entrained-channels. Reproduced when the four bands increase with p < 0.05 after correction and delta and theta do not change significantly, under one reading, with contrasts within tolerance; partially when signs and significance hold only.
- power-excited-channels
On the 78 channels near neurons whose FR rose, log power increased in all bands.
Channels near excited neurons had increases in log power across all frequency bands (emmean = 0.04, 0.08, 0.10, 0.10, 0.09, and 0.05
Results 3.7, paragraph 2; Figure 11B legend; Table S2.1How it will be judged
As power-entrained-channels. Reproduced when all six bands increase with p < 0.05 after correction under one reading and the contrasts are within tolerance; partially when signs and significance hold only.
- power-suppressed-channels
On the 47 channels near neurons whose FR fell, log power decreased in all bands.
and near suppressed neurons, decreases across all bands (emmean = -0.14, -0.19, -0.18, -0.16, -0.13, and 0.09
Results 3.7, paragraph 2; Figure 11B legend; Table S2.1How it will be judged
As power-entrained-channels. The legend prints +0.09 for high gamma in a list of decreases "across all bands"; Table S2.1 gives -9.25E-02, so -0.09 is used. Reproduced when all six bands decrease with p < 0.05 after correction under one reading and the contrasts are within tolerance; partially when signs and significance hold only.
- pac-entrained-channels
On channels near entrained neurons, order-normalized MI fell after 20-minute sACS for delta and theta to low gamma and to high gamma.
For channels near entrained neurons, there was a significant decrease in the order normalized MI observed in δ to low-γ, θ to low-γ, δ to high-γ, and θ to high-γ (emmean = -0.39, -0.21, -0.15, and -0.15
Results 3.8, paragraph 1; Figure 12B legend; Table S2.2How it will be judged
Methods 2.7, MI as in reading block M of the opening comment (M1/M2, O1/O2), with the L and B readings, 10-s windows over the 2 min before onset and after offset, eight band pairs. Reproduced when, under one reading, the four pairs decrease with p < 0.05 after correction and the alpha and beta pairs do not change significantly, with contrasts within tolerance (the SE printed in the Figure 12B legend for each response); partially when signs and significance hold only.
- pac-reduced-entrainment-channels
On channels near neurons with decreased entrainment, alpha to high-gamma order-normalized MI increased.
whereas channels near decreases in √|PPC| showed a significant increase in α to high-γ coupling (emmean = 0.22, SE = 0.10, p = *0.02)
Results 3.8, paragraph 1; Figure 12B legend; Table S2.2How it will be judged
As pac-entrained-channels; tolerance the printed SE. Reproduced when alpha to high gamma increases with p < 0.05 after correction and no other pair changes significantly, under one reading.
- pac-excited-channels
On channels near excited neurons, order-normalized MI rose for alpha and beta to low gamma and delta, alpha and beta to high gamma.
Channels near excited neurons showed increases in the order normalized MI between α to low-γ, β to low-γ, δ to high-γ, α to high-γ, and β to high-γ band pairs (emmean = 0.10, 0.10, 0.15, 0.12, and 0.09
Results 3.8, paragraph 2; Figure 12B legend; Table S2.2How it will be judged
As pac-entrained-channels. Reproduced when the five pairs increase with p < 0.05 after correction under one reading, with contrasts within tolerance; partially when signs and significance hold only.
- pac-suppressed-channels
On channels near suppressed neurons, order-normalized MI fell for seven of the eight band pairs (all but beta to low gamma).
and near suppressed neurons, decreases in δ to low-γ, θ to low-γ, α to low-γ, δ to high-γ, θ to high-γ, α to high-γ, and β to high-γ band pairs (emmean = -0.60, -0.37, -0.21, -0.54, -0.36, -0.29, and -0.06
Results 3.8, paragraph 2; Figure 12B legend; Table S2.2How it will be judged
As pac-entrained-channels. Reproduced when the seven pairs decrease with p < 0.05 after correction under one reading, with contrasts within tolerance; partially when signs and significance hold only. For beta to high gamma the legend prints -0.06 and Table S2.2 -0.60E-02, while the table's effect size (-0.152) and p (0.0105) with SE 0.0375 imply about -0.096; the value -0.06 with the SE tolerance covers -0.02 to -0.10.
Pipeline
- Paper
- done (1 of 1)
- Claims
- done
- Claims check
- done (ready)
- Attempt
- todo
- Questions
- none
- Verified
- none