file format for the preprocessed lexical data

After preprocessing step, the output files are saved under each subject folder.

/NAS_II/Projects/MVPA_Language/lexical/<subj_id>/stats_images_fmri/

<subj_id>: subject id, for example '3211'

The filename has the following structure:

filtered_<smooth><collection>_beta_tstat_matrix_<hemis>_mask_<roi>.mat

<smooth>: 's' for smoothened and '' (blank) for unsmoothened

<collection>: 'item' or 'exemplar'

<hemis>: 'lh' or 'rh'

<roi>: 'vtc2' or some other ROIs

For example,

filtered_exemplar_beta_tstat_matrix_lh_mask_vtc2.mat

filtered_sexemplar_beta_tstat_matrix_lh_mask_vtc2.mat

filtered_item_beta_tstat_matrix_rh_mask_vtc2.mat

filtered_sitem_beta_tstat_matrix_rh_mask_vtc2.mat

Exemplar file format

Our example file is

/NAS_II/Projects/MVPA_Language/lexical/3211/stats_images_fmri/filtered_sexemplar_beta_tstat_matrix_lh_mask_vtc2.mat

Let's look into the file:

>> whos   Name                        Size                Bytes  Class     Attributes    beta_animal_vs_tool       105x1                   840  double                 beta_label                  1x105               12950  cell                   beta_matrix               105x1648            1384320  double                 input_file                  1x111                 222  char                   run_id                    105x1                   840  double                 zstat_animal_vs_tool      105x1                   840  double                 zstat_label                 1x105               12950  cell                   zstat_matrix              105x1648            1384320  double

The value of measure is obtained from averaging the individual values of items.

the number of exemplar in total: 105

The number of voxels in the left-hemis of the vtc2: 1648

zstat_animal_vs_tool: the category label for each exemplar; 1-->animal and 2-->tool

zstat_label: the cell array containing the name label for each exemplar

zstat_matrix: the matrix containing the z-score of beta for every voxels in the roi

run_id: the run id number (1-4) for each exemplar

same goes for beta_*.

Item file format

Our example file is

/NAS_II/Projects/MVPA_Language/lexical/3211/stats_images_fmri/filtered_sitem_beta_tstat_matrix_lh_mask_vtc2.mat

Let's look into the file:

>> whos   Name                        Size                Bytes  Class     Attributes    beta_animal_vs_tool       412x1                  3296  double                 beta_label                  1x412               50812  cell                   beta_matrix               412x1648            5431808  double                 input_file                  1x107                 214  char                   run_id                    412x1                  3296  double                 zstat_animal_vs_tool      412x1                  3296  double                 zstat_label                 1x412               50812  cell                   zstat_matrix              412x1648            5431808  double

This is very similar to the exemplar file except that it's the measure of every single trial.

the number of items in total: 412, as we have around 3-4 items for each different exemplar

The number of éoxels in the left-hemis of the vtc2: 1648

zstat_animal_vs_tool: the category label for each exemplar; 1-->animal and 2-->tool

zstat_label: the cell array containing the name label for each exemplar

zstat_matrix: the matrix containing the z-score of beta for every voxels in the roi

run_id: the run id number (1-4) for each exemplar

same goes for beta_*.

Whole scan (not whole brain)

If you want to get all the zstat/beta of every voxel in the scan (before filtered by an ROI), the file is in the format of

filtered_<smooth><collection>_beta_tstat_matrix.mat

For example,

/NAS_II/Projects/MVPA_Language/lexical/3211/stats_images_fmri/filtered_sexemplar_beta_tstat_matrix
>> whos   Name                Size                    Bytes  Class     Attributes    beta_label          1x105                   12950  cell                   beta_matrix       105x126976            106659840  double                 run_id            105x1                       840  double                 zstat_label         1x105                   12950  cell                   zstat_matrix      105x126976            106659840  double

The number of the voxels in the whole scan: 64*64*31 = 126976