Actions

FMRI navigation

From ACL@NCU

Revision as of 07:18, 13 November 2014 by Aclexp (talk | contribs) (→‎Details of checking images)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

This page is for using AFNI on the acl_comp server to analyze the FMRI navigation experiment (starfield).

Preprocessing

  • These notes apply to 140.115.47.153

Creating alias

  • Create alias of commands (or “shortcuts”), under the home directory, use the nano command to edit a configuration file:
nano .tschrc.aclexp
  • Add to the bottom of the file:
alias <name_of_shortcut> <‘your command’>
  • Replace contents within <> to suit your need. For example:
alias cdnav ‘cd /media/DATA1/navigation’

allows you to type just “cdnav” (without the quotes) to change directory from anywhere to /media/DATA1/navigation. You will have to reinitiate another instance of terminal for the new settings to be effective.

GUI interface

  • You can initiate a graphic user’s interface (gui) of file manager by typing:
nautilus
  • Or a gui editor by typing
gedit
  • Note that when you initiate a gui in a terminal, you won’t be able to type new commands in that terminal. You can initiate another terminal from the SSH client gui.


Reconstructing rawdata into AFNI format

  • Change working directory to the navigation folder
cdnav
  • Change working directory to the script folder
cd scripts
  • edit the participants list
nano subjlist.txt
  • To 3D
python to3d_anat.py <numid>

-numid: numerical code of the participant.

python to3d_func.py <numid>

-numid: numerical code of the participant.

Creating stimulus time files

  • Change working directory to the scripts folder
cd /media/DATA1/navigation/scripts

-Or just type “cd scripts”, if you have followed settings in previous slides to change directory to /media/DATA1/navigation

  • Creat stimulus time files
python mkcond1D03.py <numid> <group id> 

-numid: numerical code of the participant (001, 002, … etc)

-group id: group code of the participant (allo or ego )

Running Preprocessing & Single-subject GLM

  • To carry out all preprocessing steps

-cutting out TR (pb00), outlier screening, normalization (pb01), realignment (pb02), spatial smoothing (pb03), and scaling (pb04)

./proc.prep <groupid> <initials+date> 

-groupid: the participant’s group (allo/ego)

-initials+date: participant’s initials followed by date (e.g. LPS141109 ).

  • To carry out single-subject (first level) GLM
./proc.glm <groupid> <numid> <initials+date> 

-groupid: the participant’s group (allo/ego)

-numid: numerical code of the participant

-initials+date: participant’s initials followed by date (e.g. LPS141109 ).

Checking & saving results

  • Change to group.allo folder
cdreva

-you need to create this alias

  • Change to group.ego folder
cdreve

-you need to create this alias

  • Change to folder of the participants you would like to see and save the results
./@ss_review_basic > review_<id> .txt

-id: name+date – LPS141109 -save results

./@ss_review_driver 

-check brain images


Details of checking files

  • Average outlier frac :

-notice if > 0.05

  • Average motion :

-notice if > 1

  • Fraction censored per run :

-notice if > 0.1


Details of checking images

  • Check whether there are spikes or ghosts from the brain image.

Spikes.png Ghosts.png

  • Check ghosts with “pb00” files from afni
  • Check spikes : change to folder of the participants
./@epi_review.<ID> 

-ID: initials+date – LPS141109

-after checking a run, press enter on command window to go through the next run and so on.

Saving to Excel

  • Save these three information to the excel worksheet ()

-Number TRs above not limit

-Number TRs above out limit

-Censor fraction