Top | Previous | Next

Running TRACULA

Remember...

For each new terminal that you open, you must do:

setenv SUBJECTS_DIR $TUTORIAL_DATA/diffusion_recons
cd $TUTORIAL_DATA/diffusion_tutorial

There are three steps in the TRACULA processing stream (which is called by the command trac-all):

• Pre-processing

• FSL's bedpostX

• Reconstructing white-matter (WM) pathways

For NMR-Center-Users Only

• Do not submit trac-all as a command in the cluster but run it on the command line. It will submit each subject listed in the Configuration File file as a seperate job to the cluster.

Below, we show you how to run these steps:.

• Warning: Running trac-all comands can take a significant amount of time. Therefore do not run the commands in this sections. All the data has already been processed for you!

Pre-processing

Some pre-processing of the diffusion image data is needed before WM pathways can be reconstructed by TRACULA. Specifically, these pre-processing steps are:

• Eddy-current correction
• Intra-subject registration (diffusion data to T1)
• Inter-subject registration (individual subject to a common space)

• Creation of cortical and white-matter masks from FreeSurfer recons

• Tensor fitting (using FSL's dtifit)

• Computing anatomical priors for WM pathways from training data

All of the above pre-processing steps can be run by passing the -prep flag to trac-all as follows:

trac-all -prep -c $TUTORIAL_DATA/diffusion_tutorial/scripts/dmrirc_single_subject

The -c flag calls your configuration file (described here) which contains your preferences for these pre-processing steps.

FSL's BedpostX

BedpostX is a program included in FSL that estimates probability distributions of the parameters of the ball-and-stick diffusion model at every voxel. The output of bedpostX (this voxel-wise probability information) is necessary for doing tractography on the subject. Passing the flag below will conveniently run bedpostX on the subjects specified in your configuration file now that all essential data was created in the pre-processing step:

trac-all -bedp -c $TUTORIAL_DATA/diffusion_tutorial/scripts/dmrirc_single_subject

Reconstructing WM Pathways

The last step is to generate the probability distributions for the set of major WM bundles you specified in the configuration file. This is done by simultaneously fitting the shape of each pathway to the results of the ball-and-stick model of diffusion from above and to the prior knowledge of the pathway anatomy given by the set of manually labeled training subjects which comes with TRACULA. To generate the WM pathways, you would run the following:

trac-all -path -c $TUTORIAL_DATA/diffusion_tutorial/scripts/dmrirc_single_subject

Now let's look at the results! Click on Next below. Top | Previous | Next