using MSMALL

[Shachar Gal]

we want to incorporate MSMALL in our processing pipeline (we acquired HCP-style data), but not quite sure where and how it fits in.

for each subject we have only one resting state scan in only one direction (AP), and three more task fMRIs (al AP as well)

this is what were doing now:

1) run the FreeSurfer scripts

2) run the fMRI Volume and Surface scripts (using MSMSulc) for each functional scan (rest and tasks alike)

3) clean with ICA-FIX

do i run MSMALL in the end of this process, which will result in a second registration for everything? 

also, seeing that i have only one scan of each kind, i should have identical values for the "fmri-names-list" and "output-fmri-name" variables?

Thanks,

Shachar

[Glasser, Matthew]

That’s right (note that MSMAll calculates the registration and dedriftandresample applies it.  How much resting state in minutes do you have?

 

Matt.

--
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/CAPdEuW79iE7x8JSQaGpLM%3D6dkZ7A%3D2%2BPCd7VZ2kWu-%2BfRg4_Zw%40mail.gmail.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.

[Shachar Gal]

10 minutes, 800 volumes

Thanks.

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/37E6FB69-4AB2-4BAB-8B4B-C2F812455DB9%40wustl.edu.

[Glasser, Matthew]

You might get better results if you used both the resting state and the task fMRI.

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/CAPdEuW7oQr%2BP4hqOuu0g3VyZsJXk7JoMkV4ELXic%2Bxw4z%3DiySw%40mail.gmail.com.

[Shachar Gal]

i'm not sure i understand what you mean by using both.

maybe i'll ask something more basic about the MSMAll script:

what happens with the inputs of "fmri-names-list" and "output-fmri-name"? the "output-fmri-name" is the one being registered and "fmri-names-list" is input for the algorithm?

for example, if i have rsfMRI_AP, tfMRI_Piano_AP and tfMRI_Motor_AP, and i want to clean tfMRI_Piano_AP, i should input:

--fmri-names-list=rsfMRI_AP@tfMRI_Piano_AP@tfMRI_Motor_AP

--output-fmri-name=tfMRI_Piano_AP

?

because in the example thats in the MSM batch script it seems like it merges all of these runs to a single file...

fMRINames="rfMRI_REST1_LR@rfMRI_REST1_RL@rfMRI_REST2_LR@rfMRI_REST2_RL"
OutfMRIName="rfMRI_REST"

i hope i'm being clear on whats confusing me

thanks,

Shachar

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/51A305FA-5D51-4498-8996-DBC6642A045D%40wustl.edu.

[Glasser, Matthew]

If you list all the runs in fmri-names-list delimited by @ you will run MSMAll using all of them in a merged file.  The final resampling occurs later in dedrift and resample, which is called separately.

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/CAPdEuW7Akfzr%2BP--S9Mj4H_LAjBbMGhQReFj5DA1mr%2B1oahQcw%40mail.gmail.com.

[Shachar Gal]

and what is the purpose of OutfMRIName? and how do i later apply this registration? using "DeDriftAndResamplePipeline.sh"?

could you maybe describe in a bit more detail the whole process of re-registering using MSMAll using the scripts in the Pipeline?

thanks again

Shachar

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/D053794A-0F9D-433D-889C-FC108F4F2D83%40wustl.edu.

[Glasser, Matthew]

That is the name of the concatenated timeseries that gets built.  The basic idea is that MSMAll uses myelin maps and fMRI to compute a functional alignment of align your subjects’ data.  Then dedrift and resample applies this alignment to all of your data, making MSMAll versions of all the data. You can then use the improved alignment in downstream cross-subject analyses.

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/CAPdEuW6yqVevq%2BaGLsXCA5T%3DTBsrpz04KHWBaFbo73CbD7cJCQ%40mail.gmail.com.

[Shachar Gal]

sorry for bothering you, but i'm still not really getting it.

for each subject, i run the MSMAll script, which creates the functional data needed for the MSMAll-based re-registration

then, i run DeDriftAndResamplePipeline for each of the subjects' functional scans (task and rest alike), which applies this re-registration?

if so, i could use some clarifications on running DeDriftAndResamplePipeline.

• which of the input variables should be the output of the MSMAll pipeline?
• which of the input variables should be the data i wish to re-register?
  
• whats the "dontFixNames" variable for?
• RegName="MSMAll_InitalReg_2_d40_WRN": is this something i need to change to match something in my data? what does the "2_d40_WRN" stand for?
• ConcatRegName="MSMAll_Test": what is the difference between RegName and ConcatRegName?
  
• DeDriftRegFiles  - is this also something i need to change to match something in my data? or is this template good for me as well?
  

thanks,

Shachar

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/A2296EA0-8DFB-47EE-B36D-9802FA391E89%40wustl.edu.

[Harms, Michael]

 

Hi,

I’m not sure if there is a misconception here, but MSMAll is not “cleaning” the data.   --fmri-names-list is the list of input fMRI runs that you want the MSMAll registration to use as part of calculating the registration.

 

Cheers,

-MH

 

-- 

Michael Harms, Ph.D.

-----------------------------------------------------------

Associate Professor of Psychiatry

Washington University School of Medicine

Department of Psychiatry, Box 8134

660 South Euclid Ave.                        Tel: 314-747-6173

St. Louis, MO  63110                          Email: mha...@wustl.edu

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/CAPdEuW7Akfzr%2BP--S9Mj4H_LAjBbMGhQReFj5DA1mr%2B1oahQcw%40mail.gmail.com.

[Glasser, Matthew]

For DedriftAndResample, it is all about what you want done to the data.  If you want FIX reapplied to fMRI data (you do if you have already cleaned that data with FIX) then add it to the fix names (and these days probably you wouldn’t use the dontFixNames, as we recommend cleaning all fMRI data with FIX). 

 

The RegName will need to match the output of MSMAll, but you can set ConcatRegName to whatever you want.  I think you can keep the DeDriftRegFiles unless Tim disagrees. 

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/CAPdEuW7zMgoBCDkq4MutqVMnHzecYsCQRPdTuZZeoLWA_d1bxg%40mail.gmail.com.

[Shachar Gal]

thanks Michael,

i see now that i referred in one of the previous replies to "cleaning", when i meant re-registering.

it was by mistake.

Shachar

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/FEFB84A9-19BE-42BF-A48B-25D80E107AE7%40wustl.edu.

[Shachar Gal]

thanks Matt,

so i should set fixNames to all of the images of this subject that i want to re-register and re-apply fix? why actually do we re-apply fix? 

and could you also specify which input variable should be the set as the output of the MSMAll pipeline?

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/A1233EF2-0E27-4E79-BDDF-5FC3C951A659%40wustl.edu.

[Glasser, Matthew]

The RegName is what references the output of MSMAll.  We reapply FIX because we start from the native mesh timeseries when resampling and that hasn’t been cleaned.

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/CAPdEuW7Qp7AQ5TpqD42OfQ%2BX-MMA3Q5GzRCcboX8_ZfxhB1z_Q%40mail.gmail.com.

[Shachar Gal]

alright

i hope it'll work

i'll keep you posted.

many thanks for your help.

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/55DFC0E9-7365-46E7-AD36-508C367D8531%40wustl.edu.

[Timothy Coalson]

I don't know exactly what Matt did when making the template data for MSMAll.  Drift should *never* be recomputed as part of registering to an existing template, but I don't know exactly how the script uses the dedrift options.

Tim

To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/A1233EF2-0E27-4E79-BDDF-5FC3C951A659%40wustl.edu.