The frequency gradient of human resting-state brain oscillations follows cortical hierarchies

  1. Keyvan Mahjoory  Is a corresponding author
  2. Jan-Mathijs Schoffelen
  3. Anne Keitel
  4. Joachim Gross  Is a corresponding author
  1. Institute for Biomagnetism and Biosignalanalysis (IBB), University of Muenster, Germany
  2. Radboud University Nijmegen, Donders Institute for Brain, Cognition and Behaviour, Netherlands
  3. Psychology, University of Dundee, Scrymgeour Building, United Kingdom
  4. Centre for Cognitive Neuroimaging (CCNi), University of Glasgow, United Kingdom
  5. Otto-Creutzfeldt-Center for Cognitive and Behavioral Neuroscience, University of Muenster, Germany
6 figures and 1 additional file

Figures

Figure 1 with 2 supplements
Spatial gradient of peak frequency (PF) across the human cortex follows the posterior-anterior hierarchy.

(A) Estimating the power spectrum for each cortical region, and identifying peak frequencies after fitting and subtracting the arrhythmic 1/f component. (B) Top-left panel shows the distribution of …

Figure 1—figure supplement 1
Stability of the PF gradient over time PF gradient along the posterior-anterior direction computed for 1 st and 2nd halves after splitting the time course to two equal segments.

(A) Top panels: correlation between trimmed mean PF (187 participants, 384 ROIs) and ROI’s location along the y-axis (posterior to anterior) computed for the first (r = −0.82, p<0.001) and second (r …

Figure 1—figure supplement 2
Within- and between-participant variability of ROIs’ size and their impact on PF gradient.

(A) Histogram of ROIs’ size for a given participant. (B) Bar plots depicting the between-participant variability of the ROIs’ area. Bars represent the mean and error bars show the standard deviation …

Spatial distribution of 1/f components (offset and slope) across human cortex.

(A) Top panel: t-values obtained from linear mixed effect modeling of 1/f offset as a function of the coordinates of ROI centroids. Bottom panel: cortical map of the corresponding fixed effects. (B) …

Figure 3 with 2 supplements
Spatial gradient of cortical thickness along the posterior-to-anterior direction.

Top panel: correlation between mean cortical thickness and ROI’s location along the y-axis (posterior to anterior) (r = −0.84, p<<0.001). Bottom panel: cortical map of trimmed mean PF across 384 …

Figure 3—figure supplement 1
Relationship between PF (187 participants, 384 ROIs) and CT, after regressing out the effect of ROI coordinates.

We factored out the impact of ROI coordinates (x,y,z) from both PF and CT using LMEM according to Equation 1, and obtained the residuals, PFres and CTres. These residuals describe individual spatial …

Figure 3—figure supplement 2
Relationship between PF (187 participants, 384 ROIs) and CT.

The scatter plot represents the dependency between the trimmed mean PF and the trimmed mean-CT and their correlation (r = −0.14, p<0.001). This low correlation (although very significant) actually …

Figure 4 with 2 supplements
PF and CT variation along the cortex follows anatomical hierarchies.

(A) Top panel: Schematic representation of seven regions (V1, V2, V4, MT, DP, TEO, 7A) used for defining visual hierarchy. Bottom panel: Each bar shows the fixed effect of the LMEM where the PF/CT …

Figure 4—figure supplement 1
Distribution of the location independent PF (PFres) among resting state networks.

T-values obtained from linear mixed effect modeling of the location independent PF among resting state networks. The network variable was defined as described in Figure 4. Here, we asked the …

Figure 4—figure supplement 2
Distribution of location independent CT (CTres) among resting state networks.

Similar to Figure 4—figure supplement 1, LMEM was performend on CT values, and found a significant difference between network. This result turns out that the significant impact of networks on CT is …

Histogram of spectral peaks.

Histogram of all detected spectral peaks (across ROIs and participants) delineates the classical frequency bands used in the EEG and MEG literature (theta 3.5–7.5 Hz, alpha 8.5–13 Hz, low-beta 15–25 …

Figure 6 with 2 supplements
Spatial gradients of band-specific PFs across human cortex follows the posterior-anterior direction.

(A, B, and C) Top panel: Dependency between the trimmed mean of band-specific PF (187 participants, 384 ROIs) and the ROI’s location along the y-axis (posterior to anterior) for theta (r = 0.4, p<0.0…

Figure 6—figure supplement 1
Spatial gradients of band-specific PF across the cortex after factoring out the impact of 1/f components (offset and slope).

(A) Top panel: Scatter plots representing the relationship between trimmed mean PFres (187 participants, 384 ROIs) and ROI’s location along the y-axis for theta band (r = 0.3, p<0.001), together …

Figure 6—figure supplement 2
Dependency between band-specific PF and CT.

Scatter plots show the relationship between trimmed mean band-specific PF and mean CT for theta (A), alpha (B), and beta (C) frequency ranges. Robust correlation was computed between trimmed mean …

Additional files

Download links