Release Name: 202108_01

Thank you for your interest in our study.

** 202108_01 release is a correction of the 202101_01 release. In the 202101_01 release,
the original files were erroneously compressed in a way that the file structures of the
original files were preserved in the compressed files. We've therefore recompressed and
uploaded as a new data release here. Please note that no other changes were made. We
apologize for any inconvenience.

The 202108_01 release of the CERB dataset package consists of two files.

1. fMRI_prepro.tar.gz
a. This file contains a folder named data_allrun. The folder structure of the data_allrun
is as follows:
i. data_allrun >> [dates of imaging session] >> [run#] >> data
ii. For example, a file named pb04.20190703jp.r01.blur+tlrc.nii can be found under
the folder structure [data_allrun >> 0703 >> run1 >>
pb04.20190703jp.r01.blur+tlrc.nii], and one can surmise that the resting state
data was acquired on 2019/07/03, and is one of the two runs of data acquired on
that date.
b. The fMRI data was preprocessed using the Analysis of Functional NeuroImages
(AFNI) software (version AFNI_20.1.06). The preprocessing pipeline included:
1) despiking, 2) slice timing correction, 3) motion correction, 4) co-registration,
5) normalization, 6) segmentation, and 7) spatial smoothing using a 6 mm (i.e., twice
the nominal acquisition voxel size) full-width at half-maximum Gaussian kernel.

2. EGGGICA_prepro.tar.gz
a. This file contains a MATLAB .mat file named EGGGICA_prepro_ds_all, in a structure
array format. The structure contains the following fields:
RSN component numbers: Our of the 42 ICA components, 18 were identified as
resting state networks (RSNs). This array contains the component numbers of the
18 RSNs
Corresponding network names: Network names of the 18 RSNs
Re-ordering information: You can use this to reorder the RSNs that is pleasing
to eyes (e.g., visual networks are group together, etc.)
EGG: preprocessed EGG. Size: [signalLength x 1]
GICAtc: time courses of RSNs. Size: [#dynamics x #RSN]
GICAic: spatial maps of RSNs. Size [#voxels x #RSN]
b. For example, if you wanted to look up the preprocessed EGG data that was acquired
on day 1, run 2, you would use EGGGICA_prepro_ds_all.sess(1).run(2).EGG
c. EGG data was preprocessed following the pipeline developed by Rebollo, et al.*,
using the FieldTrip toolbox (**, Matlab (Natick, MA;
version R2018a), and custom code provided by Rebollo, et al.
( Data were low-pass filtered
Below 5 Hz to avoid aliasing of higher-frequency signals, e.g., cardiac, and downsampled to
10 Hz. To identify the EGG peak frequency (0.033–0.066 Hz) for each run, we computed
the spectral density estimate for each EGG channel over the 900 s of EGG signal acquired
during each fMRI scan using Welch's method on 200 s time windows with 150 s overlap.
For each run, the spectral peak was identified by looking for a sharp peak within the
normogastric frequency range of 0.033–0.066 Hz. Data from the EGG channel with the
highest spectral peak were then bandpass filtered to isolate the signal related to gastric
basal rhythm (linear phase finite impulse response filter, FIR, designed with Matlab
function FIR2, centered at EGG peaking frequency, filter width ±0.015 Hz, filter order of
5). Data were filtered in the forward and backward directions to avoid phase distortions
and then further downsampled to match the sampling rate of the BOLD acquisition
(0.5 Hz).

*Rebollo I, Devauchelle AD, B ́eranger B, Tallon-Baudry C. Stomach-
brainsynchrony reveals a novel, delayed-connectivity resting-state network in
humans.Elife. 2018;7:e33321.
**Oostenveld R, Fries P, Maris E, Schoffelen JM. FieldTrip: open source software
for advanced analysis of MEG, EEG, and invasive electrophysiological data.
Computational intelligence and neuroscience. 2011;2011:1