Claims 22
From the paper Electroencephalogram signatures of loss and recovery of consciousness from propofol · claims file · checked 2026-10-22: ready
- fig1c-clicks-lost-before-verbal
In all 10 subjects, during induction, the probability of response to clicks was below that to verbal stimuli in the transition between EON and LOC.
During induction, in 10 of 10 subjects, we observed that Pclicks < Pverbal during the transition between EON and LOC
Results, Continuous Transitions Between Consciousness and Unconsciousness, paragraph 2; Figure 1CHow it will be judged
Common method, with Open A and B. Per subject, Pclicks < Pverbal holds where the posterior Pr{Pclicks < Pverbal} exceeds 0.95, which is where the 95% credibility interval of Pverbal - Pclicks excludes 0 (Results, paragraph 2). The text does not say over how much of the EON to LOC interval this must hold. Readings, (i) at some time in the interval; (ii) over more than half of it. The subject count is compared with the expected value.
- eon-loc-duration
The transition from EON to LOC took a median of 23.9 min across subjects (range 6.7 to 32.8 min, quartiles 16.4 and 28.6 min).
The transition between EON and LOC required a median of 23.9 min (minimum, 6.7 min; 25th percentile, 16.4 min; 75th percentile, 28.6 min; maximum 32.8 min)
Results, Continuous Transitions Between Consciousness and Unconsciousness, paragraph 2How it will be judged
EON and LOC per subject by the common method, with Open A and B. Reproduced when the median is within tolerance under a reading; partially reproduced when the median is and fewer than three of the other four values are. If the dataset holds the paper's LOC and EON times, they are used as well and reported as a variant.
- fig1e-group-induction
At the group level, aligned at LOC, Pclicks was below Pverbal during the transition to LOC, and the difference grew as subjects approached LOC.
The group-level analysis corroborated the individual subject analyses, showing that Pclicks < Pverbal during the transition to LOC (Pr{Pclicks < Pverbal} > 0.95; Fig. 1E). Moreover, the group-level curves showed a gradual increase in the difference between these response probabilities as subjects approached LOC
Results, Continuous Transitions Between Consciousness and Unconsciousness, paragraph 2; Figure 1 D and EHow it will be judged
Group behavior by the common method, with Open A and B. Reproduced when Pr{Pclicks < Pverbal} > 0.95 at the group level before LOC, within the window from the median EON (about 24 min before LOC) to LOC, and the posterior estimate of Pverbal - Pclicks is higher in the last third of that window than in its first third. The window and the thirds are a reading of "during the transition" and "gradual increase", named here before the analysis. Partially reproduced when only the first part holds.
- fig1e-group-emergence
During emergence, at the group level aligned at ROC, Pclicks was below Pverbal, while individual curves were more variable.
During emergence, individual response-probability curves were more variable, but again we found that Pclicks < Pverbal at the group level
Results, Continuous Transitions Between Consciousness and Unconsciousness, paragraph 3; Figure 1EHow it will be judged
Group behavior by the common method, aligned at ROC, with Open A and B. Reproduced when Pr{Pclicks < Pverbal} > 0.95 at some time after ROC, within the window from ROC to the median ROC to EOFF time (about 25 min).
- eoff-not-reached
During emergence 6 of 10 subjects did not reach EOFF, as Pclicks did not stay above 0.95 for 5 min or longer.
During emergence, 6 of 10 subjects did not reach EOFF because they did not achieve Pclicks > 0.95 for 5 min or longer
Results, Continuous Transitions Between Consciousness and Unconsciousness, paragraph 3How it will be judged
EOFF per subject by the common method, with Open A and B. The tolerance is one subject, as 10% of 6 is less than one subject and a count can only move by whole subjects.
- roc-eoff-duration
The transition from ROC to EOFF (or the end of the recording when EOFF was not reached) took a median of 25.5 min (range 4.9 to 47.6 min, quartiles 12.8 and 38.0 min).
We found that the transition between ROC and EOFF required a median of 25.5 min (minimum, 4.9 min; 25th percentile, 12.8 min; 75th percentile, 38.0 min; maximum, 47.6 min)
Results, Continuous Transitions Between Consciousness and Unconsciousness, paragraph 3How it will be judged
ROC and EOFF per subject by the common method, with Open A and B, the end of the recording standing for EOFF when it is not reached. Reproduced when the median is within tolerance under a reading; partially reproduced when the median is and fewer than three of the other four values are.
- fig2bc-gamma-beta-before-loc
Gamma and beta power rose significantly above baseline in the 30 min before LOC and stayed elevated during unconsciousness.
gamma (25–40 Hz) and beta (13–24Hz) power increased significantly above baseline levels during the 30 min before LOC ... and remained elevated during the unconscious period
Results, Dynamics of the EEG Spectrum Covary with Changes in Probability of Response and LOC/ROC, paragraph 1; Figure 2 B and CHow it will be judged
Common method, frontal channel near Fz, Laplacian, with Open C to G. For each band, a time bin counts as significant when the sign test of the subjects' baseline-normalized power has p < 0.05, with power (i) averaged over the band's frequencies (Fig. 2C) or (ii) taken per frequency as in Fig. 2B, the bin counting when more than half of the band's frequencies are significant. The paper does not say what share of the window must be significant; readings (a) more than half the time bins of LOC - 30 min to LOC, and (b) every bin of the last 10 min before LOC. Both band definitions (Results; Fig. 2 legend) are run. Reproduced when, under a reading, gamma and beta both pass in the 30 min before LOC and are significant in more than half the bins of LOC to LOC + 30 min (the unconscious period, as the 10 to 40 min after LOC of claim fig5-peak-max lies within it); partially reproduced when one band does.
- fig2bc-alpha-slow-at-loc
At LOC, alpha (8 to 12 Hz) and low-frequency (0.1 to 1 Hz) power both increased significantly.
At LOC, both alpha (8–12 Hz) and low-frequency (0.1–1 Hz) power increased significantly
Results, Dynamics of the EEG Spectrum Covary with Changes in Probability of Response and LOC/ROC, paragraph 1; Figure 2 B and CHow it will be judged
As in fig2bc-gamma-beta-before-loc, readings (i) and (ii) of the measure. "At LOC" is not given a window. Readings, (a) the band is significant in more than half the bins from LOC to LOC + 5 min (5 min being the paper's equilibration interval for the markers) and not in more than half the bins from LOC - 10 min to LOC - 5 min; (b) the band is significant in more than half the bins from LOC to LOC + 5 min. Reproduced when both bands pass under a reading; partially reproduced when one does.
- fig2d-spatial-loc15
Fifteen minutes after LOC, the alpha power increase was concentrated in frontal channels and the low-frequency increase was spread broadly over temporal and parietal channels.
Fifteen minutes after LOC, the increases in alpha power were concentrated in frontal channels, whereas those for low-frequency power were distributed broadly across temporal and parietal channels
Results, Dynamics of the EEG Spectrum Covary with Changes in Probability of Response and LOC/ROC, paragraph 1; Figure 2DHow it will be judged
Common method, all kept channels, Laplacian, Open C to F. At LOC + 15 min, per subject and channel, baseline-normalized band power; the group map is the median across subjects (Methods, Spectral Analysis, group scalp maps). The paper gives no channel groups; they are read from the 10-20 labels, frontal Fp, AF, F (not FT, FC), temporal T, TP, FT, parietal P, CP (not PO), occipital O, PO. Reproduced when the frontal group has the highest median alpha increase of the groups, and the low-frequency increase is above 1 in more than half of the temporal and parietal channels and its frontal median is not more than twice the temporal and parietal median. The second part is a reading of "distributed broadly"; it is named so here, before the analysis.
- fig2-emergence-reverse
During emergence the changes reversed with a similar time course, frontal alpha and low-frequency power decreasing at ROC while gamma and beta power stayed elevated after ROC.
During emergence, these changes in power occurred in reverse, following a similar time course: Frontal alpha and low-frequency power decreased at ROC, whereas gamma/beta power remained elevated throughout the post-ROC period when Pclicks < Pverbal
Results, Dynamics of the EEG Spectrum Covary with Changes in Probability of Response and LOC/ROC, paragraph 1; Figure 2 B and CHow it will be judged
As in fig2bc-alpha-slow-at-loc mirrored at ROC. Alpha and low-frequency power are significant in more than half the bins from ROC - 5 min to ROC and in fewer than half from ROC to ROC + 5 min; gamma and beta significant in more than half the bins from ROC to ROC + the median ROC to EOFF time (claim roc-eoff-duration). Readings (i) and (ii) of the measure and both band definitions. Reproduced when all four bands pass under a reading; partially reproduced when two or three do.
- slow-power-tenfold
Low-frequency power increased more than 10-fold in the unconscious state compared with baseline.
low-frequency power increased by more than 10-fold in the unconscious state compared with baseline
Results, Changes in Spatially Coherent Alpha Activity Mark Unconsciousness, paragraph 2How it will be judged
Common method, Laplacian, Open C to F. The paper cites Fig. 2 and does not say which channel or time. Readings, (a) the frontal channel near Fz at the Unconscious midpoint; (b) the same channel at LOC + 15 min (Fig. 2D); (c) the median over kept channels at LOC + 15 min. Group value is the median across subjects of mean 0.1 to 1 Hz power over mean baseline power. Reproduced when the ratio exceeds 9 (10 less 10%) under a reading.
- fig3a-induction-traveling-peak
In the transition before LOC, the median frequency of the frontal spectrum fell from 23.1 to 12.0 Hz and its bandwidth (interquartile range) from 17.4 to 9.1 Hz.
During induction, in the transition period before LOC, the median frequency decreased from 23.1 to 12.0 Hz, and the bandwidth decreased from 17.4 to 9.1 Hz
Results, Dynamics of the EEG Spectrum Covary with Changes in Probability of Response and LOC/ROC, paragraph 3; Figure 3AHow it will be judged
Methods, Spectral Analysis, paragraph 2. Group-level spectrogram of the frontal channel near Fz, aligned at LOC, baseline- and total-power-normalized; at each time the 25th, 50th and 75th percentiles of the 2 to 40 Hz power distribution; the bandwidth is the 75th minus the 25th; each series smoothed with a random-walk fixed-interval smoother whose observation variance is the sample variance of the series' first 10 s and whose state variance is the square of (yN - y1)/N (the equation is lost in the text; read from "the square of the end-to-end rate of change"). Open details. "Total power-normalized" is read as the baseline-normalized spectrum (linear ratio) divided by its sum over 2 to 40 Hz, so the percentiles are those of its cumulative sum; the group median is taken (a) before or (b) after this normalization. The start of the transition is not given; readings (i) the median EON time before LOC (claim eon-loc-duration), (ii) LOC - 30 min (Fig. 2, "the 30 min before LOC"). The end value is at LOC. Open C to F apply. Reproduced when all four values are within tolerance under a reading; partially reproduced when the median frequency falls by at least half the reported drop and the bandwidth falls.
- fig3a-emergence-traveling-peak
In the transition after ROC, the median frequency rose from 11.8 to 21.9 Hz and the bandwidth from 9.9 to 12.8 Hz.
During emergence, in the transition period after ROC, the median frequency increased from 11.8 to 21.9 Hz, and the bandwidth increased from 9.9 to 12.8 Hz
Results, Dynamics of the EEG Spectrum Covary with Changes in Probability of Response and LOC/ROC, paragraph 3; Figure 3AHow it will be judged
As fig3a-induction-traveling-peak, aligned at ROC. The start value is at ROC; the end is not given, readings (i) ROC + the median ROC to EOFF time (claim roc-eoff-duration), (ii) ROC + 30 min. Reproduced when all four values are within tolerance under a reading; partially reproduced when the median frequency rises by at least half the reported rise and the bandwidth rises.
- fig3b-frontal-median-frequency-power
Power at the median frequency had a frontal distribution at each behavioral time point.
We found that the power at the median frequency had a frontal distribution across the different behavioral time points
Results, Dynamics of the EEG Spectrum Covary with Changes in Probability of Response and LOC/ROC, paragraph 4; Figure 3BHow it will be judged
Per subject and channel, power at the group median frequency of claim fig3a-induction-traveling-peak (or its emergence counterpart) at each time point EON, Pre-LOC, LOC, Unconscious, ROC, Post-LOC, EOFF; group map the median across subjects (Results, paragraph 3). The text says the analysis is "for each subject" and does not say whose median frequency is used; readings (x) the group median frequency at the time point's offset from LOC or ROC, (y) the subject's own median frequency, computed as in fig3a-induction-traveling-peak on the subject's spectrogram. Whether power is raw or baseline-normalized is not given; readings (a) power in dB, (b) baseline-normalized. Channel groups as in fig2d-spatial-loc15 (frontal, temporal, parietal, occipital). Reproduced when the frontal group has the highest median of the four groups at every time point under a reading; partially reproduced when it does at five or six of the seven.
- figs2-band-spatial-shift
Gamma and beta power were frontal from Pre-LOC through LOC; during unconsciousness the frontal power moved into the beta and alpha bands; during emergence it shifted back toward beta and gamma, and frontal gamma power persisted through EOFF.
Gamma and beta power showed a frontal distribution from Pre-LOC through LOC; during unconsciousness the frontal power distribution moved to the beta and alpha bands ... In addition, frontal gamma power appeared to persist through EOFF
Results, Dynamics of the EEG Spectrum Covary with Changes in Probability of Response and LOC/ROC, paragraph 4; Figure S2How it will be judged
Group maps of band power (in dB, as Fig. S2 shows) at the seven time points as in fig3b-frontal-median-frequency-power, bands of Fig. S2 (alpha 8 to 12, beta 12 to 25, gamma 25 to 35 Hz). Each band's frontal excess is the frontal group median minus the occipital group median, in dB, minus the same at baseline. Reproduced when the gamma and beta frontal excesses are positive at Pre-LOC and LOC; at Unconscious the alpha excess is above the gamma excess; at Post-ROC the gamma excess is above that at Unconscious (the shift back toward beta and gamma); and the gamma excess is positive at EOFF. Partially reproduced when two or three of the four hold. The thresholds are a reading of "frontal distribution", "moved" and "shifted", named here before the analysis.
- fig4ab-occipital-alpha-mode
At baseline the principal mode of the alpha band is spatially coherent over occipital channels; its modal projection dissipates at LOC and returns at ROC, remaining prominent after ROC.
Occipital prominence of the principal mode was present while the subjects were awake, ended at LOC, and returned, and remained prominent after ROC
Results, Changes in Spatially Coherent Alpha Activity Mark Unconsciousness, paragraph 2; Figure 4 A and BHow it will be judged
Methods, Eigenvalue and Modal Projection Analyses. Per subject, multitaper cross-spectral matrices (parameters as for spectrograms) in nonoverlapping 2-s windows over the full baseline (Open C); the median over windows of the real and imaginary parts of each entry; the first eigenvector u1(f) is the principal mode. Modal projection in nonoverlapping 20-s windows, the power within u1 over total power (equation lost; read as u1^H S u1 over trace S, with S the window's cross-spectral matrix). Significance per Methods, a random circular shift per channel, 200 times, resampled 500 times with replacement, the 95% limit averaged across subjects and compared with the mean modal projection. Occipital means the channel group of fig2d-spatial-loc15 with the largest median |u1| at 10 Hz. Reproduced when the baseline mode at 10 Hz is largest over occipital channels, and the group mean modal projection of the baseline mode at 8 to 12 Hz is significant in more than half of the windows in the 10 min before LOC and after ROC + 10 min, and in fewer than half from LOC + 10 min to ROC - 10 min. The 10-min windows are a reading of "ended at LOC" and "returned" (the paper gives no window), named here before the analysis. Partially reproduced when the mode is occipital but the timing fails. Open E and F apply.
- fig4cd-frontal-alpha-mode
During unconsciousness the principal alpha mode (about 10 Hz) is spatially coherent over frontal channels; its modal projection begins after LOC and ceases at ROC.
Activity within this frontal alpha mode, characterized by the modal projection, begins after LOC and ceases at ROC ... during unconsciousness, the principal mode in the alpha range was prominent in a frontal distribution
Results, Changes in Spatially Coherent Alpha Activity Mark Unconsciousness, paragraph 2; Figure 4 C and DHow it will be judged
As fig4ab-occipital-alpha-mode, with the unconscious cross-spectral matrix. The Methods' "segments of at least 5 min extending from LOC to ROC" allows readings (a) the whole LOC to ROC period, (b) the 5 min after LOC. Reproduced when the unconscious mode at 10 Hz is largest over frontal channels, and its group mean modal projection at 8 to 12 Hz is significant in more than half the windows from LOC + 10 min to ROC - 10 min and in fewer than half of the windows in the 10 min before LOC and after ROC + 10 min, under a reading. Partially reproduced when the mode is frontal but the timing fails.
- fig4-no-low-frequency-mode
Low frequencies showed no prominent principal mode, although low-frequency power rose more than 10-fold in unconsciousness.
In contrast, low frequencies did not show a prominent principal mode
Results, Changes in Spatially Coherent Alpha Activity Mark Unconsciousness, paragraph 2How it will be judged
A claim of absence. The test is the paper's permutation test of the modal projection (fig4cd-frontal-alpha-mode, both readings of the segment), on the unconscious mode at 0.1 to 1 Hz. Reproduced when the group mean modal projection is not significant in more than half the windows from LOC to ROC; not reproduced when it is. The modal projection values at 0.1 to 1 Hz and at 8 to 12 Hz are reported either way. The 2-s windows resolve 0.5 Hz steps, so the low band holds the 0.5 and 1 Hz bins only; this is the paper's method and is said so.
- fig5-trough-max-induction
From about 20 min before LOC to about 10 min after it, alpha/beta (8 to 14 Hz) amplitude was largest at the troughs of the low-frequency (0.1 to 1 Hz) oscillation (trough-max), significant in 9 of 10 subjects within 10 min of LOC.
During induction, beginning ∼20 min before LOC and extending ∼10 min after LOC, the alpha/beta amplitudes were largest at the troughs of low-frequency oscillations (Fig. 5 B and C; P < 0.05 in 9 of 10 subjects within 10 min of LOC by permutation test
Results, Different Patterns of Phase-Amplitude Modulation, paragraph 2; Figure 5 B and CHow it will be judged
Methods, Phase-Amplitude Modulation Analysis, and SI. Frontal channel near Fz, Laplacian, data at 250 Hz; bandpass 0.1 to 1 Hz and 8 to 14 Hz; Hilbert phase and amplitude; 18 equal phase bins; per 2-min epoch, mean amplitude per bin over the epoch's mean. Phase 0 is the low-frequency peak (SI, "psi(t) = 0"), so trough-max is a modulogram whose largest bin has |phase| > pi/2. Open, filter design (ref. 36), readings (a) zero-phase FIR, (b) zero-phase Butterworth. Open, the permutation test (ref. 36 only), readings (a) amplitude series circularly shifted by a random offset within the epoch, statistic the modulation index of ref. 35 (Tort et al.), (b) the same shuffle, statistic the modulogram's largest bin; 200 shuffles, p < 0.05. Open, "within 10 min of LOC", readings (i) any 2-min epoch from LOC - 10 to LOC + 10 min significant and trough-max, (ii) the pooled 20 min. The tolerance is one subject (10% of 9 is less than one subject). Reproduced when the count is within tolerance under a reading, and the group modulogram (mean across subjects) is trough-max in more than half the epochs from LOC - 20 to LOC + 10 min. Partially reproduced when one of the two holds. Open C, E and F apply.
- fig5-peak-max
During profound unconsciousness, 10 to 40 min after LOC, the modulation shifted by 180 degrees, alpha/beta amplitude being largest at low-frequency peaks (peak-max), significant in 8 of 10 subjects.
During profound unconsciousness, i.e., from 10–40 min after LOC, the phase–amplitude modulation shifted by 180°, such that the alpha/beta amplitudes were largest at the peaks of low-frequency oscillations (Fig. 5 B and D; P < 0.05 in 8 of 10 subjects by permutation test
Results, Different Patterns of Phase-Amplitude Modulation, paragraph 2; Figure 5 B and DHow it will be judged
As fig5-trough-max-induction, with peak-max a modulogram whose largest bin has |phase| < pi/2, over LOC + 10 to LOC + 40 min, readings (i) any epoch and (ii) the pooled 30 min. Reproduced when the count is within tolerance under a reading and the group modulogram is peak-max in more than half the epochs of that window; partially reproduced when one of the two holds.
- fig5-trough-max-emergence
During emergence, the modulation reverted to trough-max about 15 min before ROC, significant in 9 of 10 subjects.
During emergence, the phase–amplitude modulation reverted to the trough-max pattern at ∼15 min before ROC (Fig. 5B; P < 0.05 in 9 of 10 subjects, permutation test)
Results, Different Patterns of Phase-Amplitude Modulation, paragraph 2; Figure 5BHow it will be judged
As fig5-trough-max-induction, aligned at ROC. The window is not given; readings (i) any epoch from ROC - 15 min to ROC, (ii) the pooled 15 min. Reproduced when the count is within tolerance under a reading and the group modulogram turns from peak-max to trough-max between ROC - 25 and ROC - 5 min (about 15 min, widened by 10 min either way, named here before the analysis); partially reproduced when one of the two holds.
- figs3-broadband-modulation
The low-frequency phase modulates a broad range of higher frequencies, from theta through gamma, in both the trough-max and the peak-max regimes.
These results show that the low-frequency oscillation modulates a broad range of higher frequencies spanning theta through gamma bands in both the trough-max and peak-max regimes
Figure S3; SI Materials and Methods; Results, Different Patterns of Phase-Amplitude Modulation, paragraph 2How it will be judged
SI Materials and Methods. Peaks of the low-frequency oscillation where its phase is 0; spectrograms in windows of +-2 s around each peak (T = 4 s, TW = 1, K = 1, no overlap, 0.5 Hz resolution); median over windows within a 2-min epoch, normalized per frequency by the mean over lags; epochs centered at LOC (trough-max) and at LOC + 30 min (peak-max); averaged across subjects. The SI states T = 4 s windows but also +-2 s lags, so the windows are read as the 4-s window centered on each peak, with the spectrogram's time axis giving the lag; the step along the lag is open, readings (a) 0.1 s, (b) 0.25 s. The SI also says "Within each 10-min epoch" before giving the 2-min epoch; the 2-min epochs it names for LOC and LOC + 30 min are used, and a 10-min epoch centered at the same times is shown as a variant. Reproduced when, in both epochs, the normalized power in each of theta (4 to 8 Hz), alpha, beta and gamma (25 to 35 Hz) varies with lag by more than 0.1 of the mean (the colour range of Fig. S3 is -0.2 to 0.2), at lag 0 in the opposite direction for the two epochs. Partially reproduced when this holds for alpha and beta only.
Pipeline
- Paper
- done (1 of 1)
- Claims
- done
- Claims check
- done (ready)
- Attempt
- todo
- Questions
- none
- Verified
- none