A practical modification to a resting state fMRI protocol for improved characterization of cerebrovascular function
DOI: https://doi.org/10.1101/2021.02.15.431289

BREATHING TASK CODE: https://github.com/RayStick/BreathingTasks_PsychoPy

Breath_Hold.py

scan_trigger = 5; doRest = 2; tResting = 480 ; trialnum = 3; tPace = 24 ; tBreathPace = 6; tHold = 15; tExhale = 2; tRecover = 4; BH_instructions = 'BREATH-HOLD task \n \nFollow the breathing instructions \n \nBreathe through your nose'; end_exp_key = 'escape'

Cued_Deep_Breathing.py

scan_trigger = 5; doRest = 2; tResting = 480; trialnum = 2; tStartRest = 28; tGetReady = 2; tCDB = 8; tCDBPace = 4; tFree = 43; CDB_instructions = 'DEEP BREATHING task \n \nTake deep breaths IN and OUT when cued \n \nBreathe through your nose'; end_exp_key = 'escape'

Fixation.py

RestDuration = 480; end_exp_key = 'escape'

ANALYSIS CODE (THIS REPO)

Step 1 - x.PreProc_MRI

Used the default options (except for input and output directories).

Step 2 - PreProc_PHYS/preprocess_physdata.m

Ran three times for each data segment.

inputfile_txt = physiological text file from the whole scan session
time_col = 1
vol_col = 2
sf = 1000
TR = 1.2
n_TRs = 534 (BH+REST) or 513 (CDB+REST) or 400 (REST)
prefix = subjectID_BHREST or subjectID_CDBREST or subjectID_REST
extra_TRs = 17
inputfile_json = 0

Step 3 - PreProc_PHYS/calc_CO2_regressor.m

Ran three times for each data segment.

filename = output from Step 2
fs = 1000
vol_col = 1
CO2_col = 2
prefix = subjectID_BHREST or subjectID_CDBREST or subjectID_REST
HRFconv= 1
output_hires = 1
td = 0
demean = 0
peak_check = 1

Step 4 - PreProc_PHYS/Shift_hires_CO2regressor

fMRI_file = input is the output from Step 1. CO2_file = input is the output from Step 3.

Ran in five different ways to make five data segments of length 7 mins 50 seconds:

1. Filename in 'BH+REST'; filename out 'BH+REST';
   cut_TRs_start=10;
   cut_TRs_end=134;

2. Filename in 'CDB+REST'; filename out 'CDB+REST';
   cut_TRs_start=10;
   cut_TRs_end=113;

3. Filename in 'BH+REST'; filename out 'rest_bh';
   cut_TRs_start=144;
   cut_TRs_end=0;

4. Filename in 'CDB+REST'; filename out 'rest_cdb';
   cut_TRs_start=123;
   cut_TRs_end=0;

5. Filename in 'rest'; filename out 'rest';
   cut_TRs_start=10;
   cut_TRs_end=0;

Remaining inputs (for the main analysis):
TR = 1.2
extra_TRs_half = 17
sf = 1000
nf = 40
bulk_shift = 0
save_bulk_hires = 0
fine_shift = 1
save_fine_hires = 0
shift_unit_s = 0.3
shift_max_s = 15
prefix = subjectID_BHREST_BH_CO2

Also ran with bulk_shift=1 and fine_shift=0 to explore the bulk shift (Fig4).

Step 5 - GLM_CO2shifts/x.3dREMLfit_shifts

Ran for each data segment, using the same start and end time defined in Step 4.

Input 1 = Output directory of PreProc_MRI
Input 2 = Output file from PreProc_MRI
Input 3 = Output file from PreProc_MRI
Input 4 = Output file from PreProc_MRI
Input 5 = Output file from Shift_hires_CO2regressor
Input 6 = Output file from Shift_hires_CO2regressor (demeaned with 1d_tool.py)
Input 7 = 100
Input 8 = 0.3
Input 9 = 1
Inputs 10 and 11 = Same start and end times specified for the 5 segments in Step 4
Inputs 11 = 1.2

(x.3dDeconvolve_shifts was not run, but is an alternative to x.3dREMLfit_shifts)

Step 6 - GLM_CO2shifts/x.Threshold_Maps

Ran the code for each each data segment.

Input 1 = Output directory (Concat_3dREMLfit) from Step 5
Input 2 = File prefix used for the output files from Step 5
Input 3 = Output file from PreProc_MRI
Input 4 = Output file from PreProc_MRI
Input 5 = 5.1280e-04
Input 6 = 390