Main content

Home

Menu

Loading wiki pages...

View
Wiki Version:
**When working with this dataset, please refer to the accompanying article: Škoch, A., Rehák Bučková, B., Mareš, J. et al. Human brain structural connectivity matrices–ready for modelling. Sci Data 9, 486 (2022). https://doi.org/10.1038/s41597-022-01596-9** We are making available the **structural connectivity matrices** of 88 healthy subjects. Tractography was performed using ProbTrack - FDT - FMRIB's Diffusion Toolbox, and we used 90 regions of the AAL atlas to define regions of interest (AAL_regions.csv). More detailed process of our methods and analysis will be available in our Data Report (in preparation); we will update the information on this site in due course. ------------------------------------------- > structural_connectivity_matrices - contains the structural connectivity matrix of each subject - 88 *.csv files, each with a 90x90 matrix. A given number on the x,y position represents the ***PROPORTION*** of streamlines seeded in region x that reached region y. > `SCmatrices88healthy.mat` - a single file containing the same data as the individual .csv files, in a Matlab readable format > `AAL_regions.csv` - contains the numbers and labels of 90 AAL regions, in the order in which they are listed in the structural connectivity matrices ***CAVEAT**: The Dffusion Weighted MRI DOES NOT carry the information about the directionality of white matter tracts, which results in mostly (but not entirely) symmetrictal structural connectivity matrix. The assymetry is a result of tractography, and should not be interpreted as directional.* ------------------------------------------- > structural_connectivity_matrices_not_normalized - contains the structural connectivity matrix of each subject - 88 *.csv files, each with a 90x90 matrix. A given number on the x,y position represents the ***NUMBER*** of streamlines seeded in region x that reached region y. ***CAVEAT**: The Dffusion Weighted MRI DOES NOT carry the information about the directionality of white matter tracts, which results in mostly (but not entirely) symmetrictal structural connectivity matrix. The assymetry is a result of tractography, and should not be interpreted as directional.* ------------------------------------------- > scripts - contains all functions necessary to produce the structural connectivity matrices from the raw diffusion and structural data (described below): - the 4 numbered scripts should be run on all subjects (all subfolders of `data`), in that order - utils and AAL files necessary for the 4 scripts to run - The scripts require FSL. Running them should result in structural connectivity matrices (a .csv file for each subject) similar to those shared here, under `structural_connectivity_matrices`. Note that the result should be similar but is unlikely to be exactly the same (due to the probabilistic nature of the data, possibly different version of FSL, different selection of volumes to exclude, etc.). ------------------------------------------- > data >> dti_tra_epi_2D_MDDW30 - Contains diffusion data: - `S*_diff.bval` - b-values - `S*_diff.bvec` - b-vectors - `S*_diff.nii.gz` - raw diffusion data without any preprocessing - (where applicable) `S*_volumesToExclude.txt` - contains the numbers of volumes to be excluded for the given subject (deemed to contain artifacts), where applicable. Since this visual quality control is quite subjective, you may wish to examine the data yourself and adjust these files if you find different volumes to exclude. >> t1_sag_mpr - Contains structural imaging (T1) data. - To preserve our subjects' anonymity, raw structural images are not shared. We share, instead, the outputs of *BET*, *fslmaths*, and *epi_reg* (which we already applied to the raw images; see lines 92-104 and 111-120 of script `2_dti_reg`) necessary for further steps of the processing: - `S*_data2struct_fast_wmseg.nii.gz` - the white matter mask in subject T1 space - `S*_data2struct.mat` - the subject T1-DWI transformation matrix - `S*_struct_ori_crop_brain.nii.gz` - the skull-stripped T1 ------------------------------------------- > `clinics.csv` - clinical information about subjects: - **ID**: subject identifier - **SEX**: male/female - **AGE**: in years - **RIGHT HANDEDNESS**: boolean - **EDUCATION**: elementary school / high-school / bachelor's degree or higher / - **WEIGHT**: in kg - **HEIGHT**: in m - **BMI** ------------------------------------------- For more work of our **COBRA** group, please visit our [website][1] [1]: http://cobra.cs.cas.cz/