Afni: Making Pretty Images and Saving into Powerpoint
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

Making Pretty Images

Masking Functionals

1) Run 3dAutomask on the preprocessed spiral image to make the mask:

>3dAutomask study1_norm+orig

The result will be a fim image (typically 64x64xnas). Afni will name it
automask* by default and treat it as a functional image (You will use this image to mask out data floating around outside the brain).

>afni

Display the high resolution anatomical image, choose "See Function" and select your automask. It will look something like this (See Masking page for more information about masks):


2) Apply the Binary Mask to a Functional File:

>3dcalc -prefix calc_out -a study1_norm+orig -b automask+orig -expr 'a*b'

Here you are masking out bogus data floating around outside of your yellow mask. In
this expression, we multiply the original epan image (study1_norm+orig) by our
mask (automask+orig) and generate calc_out+orig. calc_out+orig is the cleaned up file that we will use for clustering and eventual presentation.

Better Clustering
You have a lot of control over your cluster size during data set clustering. You can choose more stringent (larger) values for the clustering volume. This will help if you have a lot of little clusters and you would prefer fewer large clusters. 3.8 should still be fine for the radius. (Bigger values would be problematic as they would allow non-contiguous voxels to be treated as neighbours). Think carefully about this during analysis so that your output images are what you want for presentation.

Smoothing the Activations
Caution: Before you run 3ddup on your functional BRIKs, you must
make sure that an appropriate resliced anatomical BRIK, and only that BRIK is
available as a parent in the same directory.

>3ddup -prefix study1_dup study1_block80_norm_p05_150ul+orig.

The above command runs 3ddup, naming the output study1_dup. It uses
study1_block80_norm_p05_150ul+orig (from the Finished Afni Tutorial Data area)
as the input. This creates a HEAD file only. Now start afni:

>afni

  • Click the "Define Datamode" button.
  • Choose "Write Many" in the lower right panel and select the file you just created with 3ddup (study1_dup).
  • If you choose "Anat resam mode: Cu", the resultant activations will be interpolated and have their square edges removed. (Choose "Anat resam mode: NN" [Nearest neighbour] if you want the activations to look just like the originals.)
  • This creates the BRIK to go with the HEAD. Close and restart afni and then display your original activations and the study1_dup file to make sure they look the same (or at least very similar). Note that afni seems to have some trouble refreshing the images correctly which can be a little confusing. You may want to check images several times and perhaps close and open afni or move some images to a different directory to be sure you've got what you want.

Exporting an Image from Afni

The following steps assume that you want an image in which the clustered functional image is overlaid on an appropriate anatomical image. If you want to print something else, you'll need to make different choices.

Open AFNI in the Exam area of interest.

>afni

  • Switch Anatomy Select a brain image (anat...not epan)
  • Switch Function Select a functional image, (usually clustered at p05 or p01).
  • Define Function Correlation slider must be set to ~p=.06 (In other words, to see all the clusters in your image, the correlation slider must be set to a less stringent p-value than the value you used to create the cluster. If it is set too low, then some clusters might not display). Set color display appropriately.
  • You will probably select an Axial Image. When the image comes up, choose the appropriate slice with the slider under the image. Usually you'll want the green crosshairs turned off (Xhairs in the upper left of the afni window).



  • On linux (or another unix machine with the proper underlying software) you can press the "Disp" button on the lower left of the displayed Afni image and change the output format to jpg or tiff (etc. etc.). On sgis you may be stuck with *.ppm as an export format (But try pressing "Disp" and see if you have other options).
  • Use the c (contrast) and b (brightness) arrows on the upper right of the afni image display window to change the contrast and brightness so that the background is a pleasing solid black, but the brain image still looks nice.
  • If you would like to export a montage of image, click the "Mont" button at the bottom of the afni image viewer and decide on the montage you want.
  • Choose Sav1:*. Name the slice using some appropriate strategy (e.g. the patient's initials, condition and slice). These image files are (typically for the sgi) *.pnm files and are saved to the directory in which afni was originally opened....even if you have multiple sessions running.
    SGI: Media Convert
  • Choose Find from the toolchest => Media Tools => Choose the "Media convert" program from the option icons.
  • Use the gui interface to select the input file (Choose "File" from the upper left hand corner. Choose "Select input file").
  • Under "Output Media" on the right, Choose "Image Files".
  • Set the output file type to tiff (in the 2nd column of options list).
  • Check the path and name of the output file and click "start" (If you intend to tranfer the image files to a PC running Windows, you may want to choose short names, 8 characters or fewer followed by .tif. (Unix machines may use *.tiff which your PC may or may not be able to handle). Longer names can sometimes be shortened in the transfer process, resulting in confusion and frustration for you at the other end)
    xv (Available on Buddy, Holly and Merlin)
    Make sure you are in the directory where your images are (this isn't strictly necessary, but will make life easier).

    >xv

  • Right click on the fishy image that comes up when xv starts.
  • On the interface that pops up: Choose "Load".
  • Browse for your image and select it. Close the loader (if it isn't already closed), and click "Save".
  • An interface pops up. In the upper right, use the drop down menu to browse for the image format you want and choose it (tiff).
  • Name your file in the text box at the bottom and click okay.
  • If you want to, choose "Load", then browse for and load the image you just created to make sure the process worked correctly.
Get tiff files with scp from any pc. Note: *.tiff will be shortened to *.tif under Windows. Sometimes names will be shortened to 8 characters followed by .tif. You may wish to use short names to avoid confusion. See ssh and scp.