Skip to content

Single subject (first‐level) task analysis

Michelle Voss edited this page Nov 2, 2023 · 13 revisions

By the end of this practical you should be able to:

  • create and understand a model of task-related BOLD signal in the FEAT GUI
  • understand how to test for activation differences between conditions
  • locate and view the activation maps in fsleyes

Access FastX through the remote login:
https://fastx.divms.uiowa.edu:3443/

Create model of task-related BOLD signal in FEAT:

  • In terminal, move yourself to your data folder cd ~/fmriLab/
  • Type fsl and click on FEAT FMRI analysis
  • Now instead of Preprocessing select Statistics in the drop-down menu at the top, like so: single-subject_feat-set-statistics-only


  • Then select Input is a FEAT directory and navigate to sub-001's flanker.feat directory:
featInput


  • Move to the Stats tab, first turn on Add additional confound EVs and then select the outliers.txt file in your beh folder as shown:
inputOutliers


  • Next, click on Full model setup. Set up your model using tabs for each explanatory variable (EV) like shown below:

EV1:
EV-con


EV2:
EV-neu


EV3:
EV-inc


EV4:
EV-errors


  • After setting up the EVs, stay in the Full model setup window and go to the Contrasts & F-tests tab. Set up the contrasts as below. Click Done when setup complete.
taskContrasts

  • After completing the Full model setup you will get your task model!
taskModel

  • The last tab in the FEAT GUI is Post-stats. We will leave the defaults on for now. The settings here are most relevant when we get to group-level analyses.


  • Click Go. When it's finished, the results will appear within your flanker.feat directory. Your html report will then include output of brain activation maps as previewed below.
featReport

  • Using fsleyes will allow us to view the results more interactively:
    • Use the html report to locate the directory where the activation maps are on your computer
    • Use the terminal to move yourself there: cd ~/fmriLab/flankerData_n4/sub-001/func/flanker.feat/
    • Open fsleyes through the terminal with settings for viewing FEAT output: fsleyes -ad filtered_func_data.nii.gz stats/zstat1 stats/zstat2 stats/zstat3
    • In the fsleyes menu, select View -> Layouts -> FEAT mode
    • You should see a display like below in fsleyes.
    • Clicking on the buttons with arrows in the column labeled Z Max location will move your cursor to the location of that peak in brain activation. Give it a try!


fsleyesFEAT

  • Once you've completed all the steps above for sub-001, repeat for sub-002 and sub-003. Example PDFs for their FEAT activation reports will be on our course ICON site. Before next class, you should have statistical maps for the same contrasts, in the SAME order, for all subjects.