Menu

Regress out covariates before ICA

2019-06-05
2019-06-06
  • Monica Nicoli

    Monica Nicoli - 2019-06-05

    Dear experts,
    in our neuroimaging team we are trying to understand how we can correct single-subject scans for motion parameters before entering them in GIFT and running the ICA.
    We have already tried the regress out covariates function in the toolbox REST, but once we entered those regressed images in GIFT, the ICA analysis could not start.
    We also read in the mailing list that several authors suggest to correct for motion impact after running the ICA. Has anybody tried to correct it at a single subject level, before running the ICA?

    Thank you so much.
    Monica Nicoli

     
    • Srinivas Rachakonda

      Hi Monica,

      You could correct for motion after ICA in GIFT mancova toolbox.
      Alternatively, you could also correct for motion before feeding into ICA.
      Inputs for the function are the nifti files and motion covariates in a cell
      array

      Thanks,
      Sri

      On Wed, Jun 5, 2019 at 8:18 AM Monica Nicoli mnicoli@users.sourceforge.net
      wrote:

      Dear experts,
      in our neuroimaging team we are trying to understand how we can correct
      single-subject scans for motion parameters before entering them in GIFT and
      running the ICA.
      We have already tried the regress out covariates function in the toolbox
      REST, but once we entered those regressed images in GIFT, the ICA analysis
      could not start.
      We also read in the mailing list that several authors suggest to correct
      for motion impact after running the ICA. Has anybody tried to correct it at
      a single subject level, before running the ICA?

      Thank you so much.
      Monica Nicoli


      Regress out covariates before ICA
      https://sourceforge.net/p/icatb/discussion/309910/thread/8552a85a44/?limit=25#001f


      Sent from sourceforge.net because you indicated interest in
      https://sourceforge.net/p/icatb/discussion/309910/

      To unsubscribe from further messages, please visit
      https://sourceforge.net/auth/subscriptions/

       
  • Monica Nicoli

    Monica Nicoli - 2019-06-06

    Thank you Srinivas.
    I took a look at the code, but it doesn't work because of some missing end statements. Once fixed them, it shows this error:

    Undefined function or variable 'tmp'.
    Error in icatb_rm_variance_mod/load_regress (line 101)
    R{nR} = tmp
    Error in icatb_rm_variance_mod (line 18)
    R = load_regress(rfiles)

    Could you help me to figure out what does it mean? It seems that line 101 is outside any if-statement, and also that the variable "tmp" could not be created. Attached you can find the modified code.
    Also, line 95 was commented in your original code. Is it correct to uncomment it?

    Thank you a lot for your reply!
    Monica

     
    • Srinivas Rachakonda

      Will take a look and Could you email srachakonda@gsu.edu from now on?

      On Thu, Jun 6, 2019 at 7:42 AM Monica Nicoli mnicoli@users.sourceforge.net
      wrote:

      Thank you Srinivas.
      I took a look at the code, but it doesn't work because of some missing end
      statements. Once fixed them, it shows this error:

      Undefined function or variable 'tmp'.
      Error in icatb_rm_variance_mod/load_regress (line 101)
      R{nR} = tmp
      Error in icatb_rm_variance_mod (line 18)
      R = load_regress(rfiles)

      Could you help me to figure out what does it mean? It seems that line 101
      is outside any if-statement, and also that the variable "tmp" could not be
      created. Attached you can find the modified code.
      Also, line 95 was commented in your original code. Is it correct to
      uncomment it?

      Thank you a lot for your reply!
      Monica


      Regress out covariates before ICA
      https://sourceforge.net/p/icatb/discussion/309910/thread/8552a85a44/?limit=25#dd47


      Sent from sourceforge.net because you indicated interest in
      https://sourceforge.net/p/icatb/discussion/309910/

      To unsubscribe from further messages, please visit
      https://sourceforge.net/auth/subscriptions/

       
      • Monica Nicoli

        Monica Nicoli - 2019-06-06

        Sure!

         

Log in to post a comment.