Confusion over how to setup MSMAll/DeDriftAndResample after Multi-Run FIX

[Chris Racey]

Dear HCP experts,

I'm having some confusion about how to setup the later stages of the pipeline.

Our experiment has two sessions (scanned on different days), with 4 runs of resting state in each. I believe it makes the most sense to use multi-run fix, concatenating the four runs from each session. Below I've pasted the key configuration lines of the last four scripts, and my questions are listed below each. I'm particularly confused about how to configure the MSMall and DeDriftandResample scripts, and what the final outputs would be that I use to perform parcellation and resting state analysis.

I'm trying to follow the process outlined in HCP users FAQ #10

###IcaFixProcessingBatch.sh
fMRINames="rfMRI_REST1_AP@rfMRI_REST1_PA@rfMRI_REST2_AP@rfMRI_REST2_PA%rfMRI_REST3_AP@rfMRI_REST3_PA@rfMRI_REST4_AP@rfMRI_REST4_PA"
# If you wish to run "multi-run" (concatenated) FIX, specify the names to give the concatenated output files
# In this case, all the runs included in ${fMRINames} become the input to multi-run FIX
# Otherwise, leave ConcatNames empty (in which case "single-run" FIX is executed serially on each run in ${fMRINames})
#ConcatNames=""
ConcatNames="rfMRI_REST12_AP_PA@rfMRI_REST34_AP_PA"  ## Use space (or @) to separate concatenation groups

###PostFixBatch.sh
# List of fMRI runs
# If running on output from multi-run FIX, use ConcatName as value for fMRINames
fMRINames="rfMRI_REST12_ AP_PA rfMRI_REST34_ AP_PA"
HighPass="2000"
ReUseHighPass="YES" #Use YES if running on output from multi-run FIX, otherwise use NO

The above run fine and I think they are configured correctly, I'm including them here as they influence how to set up subsequent scripts.

###MSMAllPipelineBatch.sh
# For MR FIX, set fMRINames to empty
fMRINames=""
# the original MR FIX parameter for what to concatenate
mrfixNames="rfMRI_REST1_AP@rfMRI_REST1_PA@rfMRI_REST2_AP@rfMRI_REST2_PA"
# the original MR FIX concatenated name
mrfixConcatName="rfMRI_REST12_AP_PA"
# @-separate list of runs to use
mrfixNamesToUse="rfMRI_REST1_AP@rfMRI_REST1_PA@rfMRI_REST2_AP@rfMRI_REST2_PA"
OutfMRIName="rfMRI_REST12"

I've set the above up to just process the runs from first session. Can it be configured to do all runs/sessions at once? I would want any concatenation to happen for each session and I'm not sure it would be useful to concatenate all 8 runs together into one.

If I set mrfixNames to the fMRINames variable from IcaFixProcessingBatch it fails because of the session split %

Similarly, if I set mrfixConcatName="rfMRI_REST12_AP_PA@rfMRI_REST34_AP_PA" it also fails.

I could run the script twice, once for session 1 and once for session 2, but it seems that would overwrite the output registration files in the fsaverage_LR32k and Native folders and only be correct/accurate for the second session.

If I did run it twice I would get two output folders (OutfMRIName="rfMRI_REST12" & OutfMRIName="rfMRI_REST34"). I'm not clear what the output in this folder is used for (rfMRI_REST12_Atlas_hp2000_clean_vn.dtseries.nii) relative to those in rfMRI_REST12_AP_PA? Perhaps OutfMRIName should be set to rfMRI_REST12_AP_PA?

###DeDriftAndResamplePipelineBatch.sh
HighResMesh="164"
LowResMesh="32"
RegName="MSMAll_InitalReg_2_d40_WRN"
DeDriftRegFiles="${HCPPIPEDIR}/global/templates/MSMAll/DeDriftingGroup.L.sphere.DeDriftMSMAll.164k_fs_LR.surf.gii@${HCPPIPEDIR}/global/templates/MSMAll/DeDriftingGroup.R.sphere.DeDriftMSMAll.164k_fs_LR.surf.gii"
ConcatRegName="MSMAll_Test"
Maps="sulc curvature corrThickness thickness"
MyelinMaps="MyelinMap SmoothedMyelinMap" #No _BC, this will be reapplied
MRFixConcatNames=" rfMRI_REST12_AP_PA"
MRFixNames=" rfMRI_REST1_AP@rfMRI_REST1_PA@rfMRI_REST2_AP@rfMRI_REST2_PA"
fixNames="NONE" #Space delimited list or NONE
dontFixNames="NONE"

Can the above also be set up to operate on both sessions at once?

And finally, after all scripts are complete, what is the correct concatenated output file that I should use to parcellate and perform resting state analysis?

I apologise if these questions are answered somewhere already, I've tried my best to find additional documentation. If there are some I would be grateful to be pointed in the right direction?

Many thanks,

Chris

[Coalson, Timothy Scott (S&T-Student)]

I think our recommendation is to just run MR FIX with all the data, being acquired on different days shouldn't be important (the HCP data for a subject is collected over a couple days), having more data with activations should make it easier to separate them from any per-day or other artifacts.  You should only run MSMAll once per subject, and the idea is for all of the resting state to be in a single MR FIX run, and for the MSMAll script to pull the resting state chunks out of the concatenated MR FIX file.

Tim

---

From: Chris Racey <c.ra...@gmail.com>
Sent: Tuesday, August 31, 2021 12:08 PM
To: hcp-...@humanconnectome.org <hcp-...@humanconnectome.org>
Subject: [hcp-users] Confusion over how to setup MSMAll/DeDriftAndResample after Multi-Run FIX

 

WARNING: This message has originated from an External Source. This may be a phishing expedition that can result in unauthorized access to our IT System. Please use proper judgment and caution when opening attachments, clicking links, or responding to this email.

--
You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.
To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/CABSuK1OYVvwOFhoEKrYUD%3DpTROAppSh4hDYVn%2BhO9Jv0ij%2BFZg%40mail.gmail.com.

[Chris Racey]

Hi Tim,

Many thanks for your response!

The goal of our study is to look at resting state differences between session 1 and session 2, there was an intervention in between. My concern would be running fix across both sessions may eliminate differences we care about? 

Perhaps there is a way to do the single-session concatenation after fix and msmall?

Best wishes,

Chris

[Coalson, Timothy Scott (S&T-Student)]

Ah, I see.  I think the way I would approach it is to still run it all in a single MR FIX, a single MSMAll, and then separate the pre and post runs after all the preprocessing is done.  This assumes that the functional areas haven't been reconfigured by the intervention (MSMAll uses functional data for the registration), which seems plausible, particularly without much delay between sessions.

Of course, depending on the details of the specific intervention, this might not work - if the intervention involves surgery, you might need new structural scans, and therefore the post-intervention data would be treated as if it were a new subject.

Tim

---

From: Chris Racey <c.ra...@gmail.com>
Sent: Tuesday, August 31, 2021 5:10 PM
To: HCP-Users <hcp-...@humanconnectome.org>
Cc: Coalson, Timothy Scott (S&T-Student) <tsc...@mst.edu>
Subject: Re: [hcp-users] Confusion over how to setup MSMAll/DeDriftAndResample after Multi-Run FIX

 

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/45734fb1-30ba-4293-a01f-dc02552fd438n%40humanconnectome.org.

[Glasser, Matt]

I would just run MR+FIX on the whole dataset and then MSMAll on the whole dataset.  It would not eliminate any differences to the neural BOLD (which should not be messed with by sICA+FIX anyway).

Matt.

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/SN6PR01MB4991040CAA836DEE3867CBA781CC9%40SN6PR01MB4991.prod.exchangelabs.com.

 

---

The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail.

[Chris Racey]

Many thanks Tim and Matt, I will do as you suggest. The intervention is a neuroinflammatory challenge and there's a very short duration between sessions so we wouldn't expect any structural changes.

Would you be able to advise on the mechanics of separating out the pre/post runs? Would I run a command to concatenate the MSMAll registered individual runs or split the fully concatenated 8run output in half? 

I'm not very clear on what specific file output, from the many that are produced, I should be using to conduct resting state analysis?

Best wishes,

Chris

[Glasser, Matt]

MR+FIX automatically splits out the runs (undoing the demeaning and variance normalization in the process), but you could simply use wb_command -cifti-merge to split the fully concatenated timeseries into session 1 and session 2 if you wanted to keep the demeaning and variance normalization (and not have to redo it).

 

Matt.