Afni: Analysis for Block Design
back to the main page
glossary of technical terms (unix and imaging) and links to other relevant technical information
links to information for neuroimagers and visitors to Tucson
search the CNL site
image analysis software info and links
A basic "How to" for Bob Cox's Automated Functional NeuroImaging analysis program

All examples are for the tutorial. Download the files from the analysis directory and unzip them to proceed:

  • study1_norm+orig.BRIK and .HEAD,
  • block80.1D,
  • brain+orig.BRIK and .HEAD,
  • optionally brain3d+orig.BRIK and .HEAD

If you wish to prepare images for presentation, there are some extra issues to consider and this is the stage (prior to analysis) at which you should do it. See the page on making Pretty Images. If you are just doing the tutorial, you might want to leave this till later.

Analysis: Phase 1

In this phase you apply the waver file to your data to see which areas of activation are correlated to the conditions defined by hemodynamic responses in that waver file.

Put the above files into an analysis directory and cd to it. Type:

>afni (or, after afni is open, choose Define Data Mode and Read Sess to set up multiple sessions).

  • A 'widget' window (GUI) will appear:
  • Switch Anatomy. A drop down window appears. Choose an appropriate *_norm file (functional epan file). Click it to highlight it, else you will not be able to access the graph. "Image" will show you a picture of the data.

  • "Graph" will show you intensity values over time associated with each voxel (The one in the yellow box in the center is the one your green crosshairs point at.)

  • At the bottom right of the graph window, choose FIM => Pick Ideal =>Choose and set the appropriate *.1D file. (Note: After you pick the ideal, it will appear on the graph as a separate line and have values for all points along the x-axis)

    • FIM => Compute FIM =>fico (watch green progress meter)
    • Define Function. This brings up additional widgets, a correlation bar and color bar. Select 4 colors, change orange to red and set the color change. If you leave the correlation slider at r=.5000, you'll rarely see activation, so check to make sure you have lowered it.

  • Rename first analysis: Define Data Mode=>Plugins=>Dataset Rename.
    -Your analysis will likely be at the bottom of the list with a prefix like study1_norm@1. Change the number to the ideal name (e.g. study1_norm@block80). Run and Close.
Cluster Analysis: Part 2
  • Define Function=>choose p value (under correlation slider). Record p and r.
  • This simply alters which functional activations will be visible. If this is a second clustering, recheck Switch Function to make sure it is using the correct dataset.
  • Define Data Mode =>Plugins=>3D Cluster

      Dataset: file from first analysis (e.g., study1_norm@block80).
      Radius: 3.8
      Min vol: 150
      Threshold: the r value from the correlation slider (double check it).
      Output: Condition_ideal_norm_pvalue_minvol (e.g., study1_block80_norm_p05_150ul)
      Run and Close
      Optional: Switch Anatomy, Select brain anatomy for pretty overlay.

  • You now have a clustered fico BRIK
  • If you want it to display like the image below, choose "Switch Anatomy" and select "brain" from the list.
  • Then choose "switch function" and select your study1_block80_norm_p05_150ul. As long as "see function" is also selected, you should now see an image like the one below:

After looking at your initial output, you may want to fiddle with clustering parameters to enhance images for presentation. To learn more about pretty images, go here.