C First Level .feat directory contents

This appendix does not explain every single file in .feat directories, only the most important ones.

C.1 Top Level .feat directory

  • design.fsf: Feat Setup File – saves all parameters entered into FEAT

  • design.mat: Design matrix file – created from design.fsf by feat_model

  • design.con: Design contrasts file – created from design.fsf by feat_model

  • design.png: Image file showing the design matrix

  • design_cov.png: Image file showing the covariance matrix of the design matrix

  • filtered_func_data.nii.gz: Pre-processed fMRI data (see pre-processing handout)

  • example_func.nii.gz: Example fMRI volume (taken from middle of timeseries)

  • mask.nii.gz: Brain mask used to skull-strip the fMRI data

  • mean_func.nii.gz: Average fMRI image (after pre-processing)

  • report.*: These files contain a report you can read in your web browser

  • thresh_zstat1.nii.gz: Volume containing the thresholded zstat1 using whichever method you chose

  • thresh_zstat1.vol: Contains the count of non-zero voxels in the unthresholded zstat

  • rendered_thresh_zstat1.nii.gz: example_func image with thresh_zstat1 overlaid

  • rendered_thresh_zstat1.png: Image of the rendered statistic

C.2 custom_timing_files subdirectory

This contains copies of your EV files named in order. These are copied in before the FEAT model is setup by feat_model.

  • ev1.txt: For our example, this is a copy of the face.txt EV file
  • ev2.txt: For our example, this is a copy of the scram_face.txt EV file
  • ev3.txt: For our example, this is a copy of the scene.txt EV file

C.3 reg subdirectory

This directory contains structural images (initial_highres, highres, standard) as .nii.gz files:

  • initial_highres.nii.gz: The skull-stripped FLAIR image (if used)
  • highres.nii.gz: The skull stripped T1 image
  • standard.nii.gz: Standard brain image – usually the skull stripped MNI152 image
  • standard_mask.nii.gz: Standard brain mask – usually the brain mask of the MNI152 image
  • standard_head.nii.gz: Standard head image – usually the non-skull stripped MNI152 image

as well as many transformation matrices (*.mat). For example:

  • example_func2highres.mat is the matrix which describes how to go from the example_func space to the highres space.
  • standard2example_func.mat is the matrix which describes how to go from the standard space to the example_func space.

These matrices are used by flirt, ApplyXFM and the co-ordinate transform tools to perform the transformations between the different spaces.

  • highres2standard_warp.nii.gz: the warp file which contains the non-linear warp from highres space to standard space.

This file, along with other warp files, are used by applywarp and the other co-ordinate transform tools to perform non-linear transformations to and from standard space (if the non-linear option was used).

C.4 stats subdirectory

This is the most important subdirectory in your FEAT analysis. It contains the statistical information for the individual run you have analysed. The important files are:

  • res4d.nii.gz: The residuals (model error) left in the fMRI data after the model has been fitted
  • peN.nii.gz: The parameter estimates (betas) for EV number N. In our case, we will have 4 Pes
  • copeM.nii.gz: The Contrast Of Parameter Estimates for contrast M
  • varcopeM.nii.gz: The Variance (Error) image for contrast M
  • tstatM.nii.gz: A t-statistic volume for contrast M calculated using the cope / varcope volumes
  • zstatM.nii.gz: A z-statistic volume for contrast M calculated from the t-statistic volume

All statistical images in the stats/ subdirectory are unthresholded.

The cope*.nii.gz and varcope*.nii.gz are the files upon which the data which is carried forward into the higher-level analyses is based. Note that this means that the thresholding method you use at the first-level has no influence on the higher-level analyses. The actual files carried forward into the higher-level analysis are the files in reg_standard/stats.

C.5 mc subdirectory

This directory contains the results of the motion correction:

  • rot.png: Image showing rotation as a function of time
  • trans.png Image showing translation as a function of time
  • disp.png Image showing displacement as a function of time

The prefiltered_func_data_mcf.mat directory contains one transformation matrix for each volume in the fMRI dataset describing how the volume was motion corrected.

C.6 tsplot subdirectory

This directory contains time series plots for each of the constrasts and EVs as well as containing a report (tsplot_index.html) which can be viewed in a web browser.

C.7 logs subdirectory

This directory contains log files describing what happened during the analysis. It is mainly useful when things go wrong!

C.8 hr subdirectory

NB: This directory will only exist if you have used the renderhighres command

  • background.nii.gz: This is a copy of the highres (T1) structural scan.
  • thresh_zstat1.nii.gz: Thresholded z-statistic in the highres space

C.9 reg_standard subdirectory

NB: This directory will only exist if you have used the feat directory in a group analysis or manually ran the featregapply command. It contains versions of various files converted into the standard space.

  • example_func.nii.gz: Version of example_func in the standard space
  • mask.nii.gz: Version of the fMRI brain mask in the standard space
  • mean_func.nii.gz: Version of the mean functional image in the standard space
  • reg/highres.nii.gz: Version of the highres (T1) image in the standard space
  • stats/cope1.nii.gz: COPE map for contrast 1 in the standard space
  • stats/varcope1.nii.gz: VARCOPE map for contrast 1 in the standard space

Note that these last files are the ones which will actually be carried forward into the group analysis.