Wasserthal Jakob
Neher Peter
Maier-Hein Klaus
2018-11-05
<p><strong>Overview</strong></p>
<p>This dataset contains segmentations of 72 white matter tracts obtained from 105 subjects included in the Human Connectome Project (HCP) young adult dataset (https://www.humanconnectome.org/study/hcp-young-adult). The folder names correspond to the ID of the HCP subjects. This dataset only contains the tracts. It does not contain the original DWI data. This has to be downloaded from the HCP website (it is free, but you have to register to get access).</p>
<p>The data is part of the following publication: <a href="https://doi.org/10.1016/j.neuroimage.2018.07.070">Wasserthal et al., TractSeg - Fast and accurate white matter bundle segmentation. NeuroImage (2018)</a>. If you use the data please cite the paper.</p>
<p> </p>
<p><strong>Details for generating corresponding whole brain tractograms</strong></p>
<p>The tracts were extracted semi-automatically from whole-brain tractograms. For a detailed description of the tract segmentation process please refer to the paper. The following MRtrix (http://www.mrtrix.org/) commands were used to obtain the whole-brain tractograms:</p>
<p>5ttgen fsl T1w_acpc_dc_restore_brain.nii.gz 5TT.mif -premasked<br>
dwi2response msmt_5tt Diffusion.nii.gz 5TT.mif RF_WM.txt RF_GM.txt RF_CSF.txt -voxels RF_voxels.mif -fslgrad Diffusion.bvecs Diffusion.bvals<br>
dwi2fod msmt_csd Diffusion.nii.gz RF_WM.txt WM_FODs.mif RF_GM.txt GM.mif RF_CSF.txt CSF.mif -mask nodif_brain_mask.nii.gz -fslgrad Diffusion.bvecs Diffusion.bvals<br>
tckgen -algorithm iFOD2 WM_FODs.mif output.tck -act 5TT.mif -backtrack -crop_at_gmwmi -seed_image nodif_brain_mask.nii.gz -maxlength 250 -minlength 40 -number 10M -cutoff 0.06 -maxnum 0</p>
<p>For "CA", "IFO_left", "IFO_right", "UF_left", "UF_right" we used tracking without anatomical constraints:</p>
<p>tckgen -algorithm iFOD2 WM_FODs.mif output.tck -seed_image nodif_brain_mask.nii.gz -maxlength 250 -minlength 40 -number 10M -cutoff 0.06 -maxnum 0</p>
<p>Due to their enormous size, the whole brain tractograms corresponding to the segmented tracts are not included this dataset. Please contact the author of the paper if you are interested in these tractograms.</p>
<p> </p>
<p><strong>Included tracts</strong></p>
<p>1: AF_left (Arcuate fascicle)<br>
2: AF_right<br>
3: ATR_left (Anterior Thalamic Radiation)<br>
4: ATR_right<br>
5: CA (Commissure Anterior)<br>
6: CC_1 (Rostrum)<br>
7: CC_2 (Genu)<br>
8: CC_3 (Rostral body (Premotor))<br>
9: CC_4 (Anterior midbody (Primary Motor))<br>
10: CC_5 (Posterior midbody (Primary Somatosensory))<br>
11: CC_6 (Isthmus)<br>
12: CC_7 (Splenium)<br>
13: CG_left (Cingulum left)<br>
14: CG_right <br>
15: CST_left (Corticospinal tract<br>
16: CST_right <br>
17: MLF_left (Middle longitudinal fascicle)<br>
18: MLF_right<br>
19: FPT_left (Fronto-pontine tract)<br>
20: FPT_right <br>
21: FX_left (Fornix)<br>
22: FX_right<br>
23: ICP_left (Inferior cerebellar peduncle)<br>
24: ICP_right <br>
25: IFO_left (Inferior occipito-frontal fascicle) <br>
26: IFO_right<br>
27: ILF_left (Inferior longitudinal fascicle) <br>
28: ILF_right <br>
29: MCP (Middle cerebellar peduncle)<br>
30: OR_left (Optic radiation) <br>
31: OR_right<br>
32: POPT_left (Parieto‐occipital pontine)<br>
33: POPT_right <br>
34: SCP_left (Superior cerebellar peduncle)<br>
35: SCP_right <br>
36: SLF_I_left (Superior longitudinal fascicle I)<br>
37: SLF_I_right <br>
38: SLF_II_left (Superior longitudinal fascicle II)<br>
39: SLF_II_right<br>
40: SLF_III_left (Superior longitudinal fascicle III)<br>
41: SLF_III_right <br>
42: STR_left (Superior Thalamic Radiation)<br>
43: STR_right <br>
44: UF_left (Uncinate fascicle) <br>
45: UF_right <br>
46: CC (Corpus Callosum - all)<br>
47: T_PREF_left (Thalamo-prefrontal)<br>
48: T_PREF_right <br>
49: T_PREM_left (Thalamo-premotor)<br>
50: T_PREM_right <br>
51: T_PREC_left (Thalamo-precentral)<br>
52: T_PREC_right <br>
53: T_POSTC_left (Thalamo-postcentral)<br>
54: T_POSTC_right <br>
55: T_PAR_left (Thalamo-parietal)<br>
56: T_PAR_right <br>
57: T_OCC_left (Thalamo-occipital)<br>
58: T_OCC_right <br>
59: ST_FO_left (Striato-fronto-orbital)<br>
60: ST_FO_right <br>
61: ST_PREF_left (Striato-prefrontal)<br>
62: ST_PREF_right <br>
63: ST_PREM_left (Striato-premotor)<br>
64: ST_PREM_right <br>
65: ST_PREC_left (Striato-precentral)<br>
66: ST_PREC_right <br>
67: ST_POSTC_left (Striato-postcentral)<br>
68: ST_POSTC_right<br>
69: ST_PAR_left (Striato-parietal)<br>
70: ST_PAR_right <br>
71: ST_OCC_left (Striato-occipital)<br>
72: ST_OCC_right<br>
</p>
<p><strong>Cross-validation data splits</strong></p>
<p>The following data splits were used for cross-validation in the TractSeg paper:</p>
<pre><code class="language-python">fold1 = ['992774', '991267', '987983', '984472', '983773', '979984', '978578', '965771', '965367', '959574', '958976', '957974', '951457', '932554', '930449', '922854', '917255', '912447', '910241', '907656', '904044']
fold2 = ['901442', '901139', '901038', '899885', '898176', '896879', '896778', '894673', '889579', '887373', '877269', '877168', '872764', '872158', '871964', '871762', '865363', '861456', '859671', '857263', '856766']
fold3 = ['849971', '845458', '837964', '837560', '833249', '833148', '826454', '826353', '816653', '814649', '802844', '792766', '792564', '789373', '786569', '784565', '782561', '779370', '771354', '770352', '765056']
fold4 = ['761957', '759869', '756055', '753251', '751348', '749361', '748662', '748258', '742549', '734045', '732243', '729557', '729254', '715647', '715041', '709551', '705341', '704238', '702133', '695768', '690152']
fold5 = ['687163', '685058', '683256', '680957', '679568', '677968', '673455', '672756', '665254', '654754', '645551', '644044', '638049', '627549', '623844', '622236', '620434', '613538', '601127', '599671', '599469']</code></pre>
<p>Hyperparameters were optimized using fold 1-3 for training and fold 4 for validation.</p>
<p>The final 5-fold cross-validation (results reported in the TractSeg paper) was done by always training on 3 folds, selecting the best epoch by evaluating on the fourth fold and then reporting the final results (of the model from the best epoch) on the fifth fold.</p>
<p>The pretrained TractSeg model which will automatically be used when you download TractSeg was trained on fold1+fold2+fold3.</p>
<p>Please use the same data splits to make your work comparable.</p>
<p> </p>
<p><strong>Data format</strong></p>
<p>From version 1.2.0 of this dataset onwards it uses the newest trackvis (trk) standard (using nibabel.streamlines API). Streamlines are saved in native voxel space and when loaded are transformed to coordinate space using the affine stored in the trk file header. In the previous versions of the dataset the older nibabel.trackvis API was used (streamlines are saved in real coordinate space and no affine is applied when loading them).</p>
https://doi.org/10.5281/zenodo.1477956
oai:zenodo.org:1477956
Zenodo
https://doi.org/10.1016/j.neuroimage.2018.07.070
https://zenodo.org/communities/dmri
https://doi.org/10.5281/zenodo.1088277
info:eu-repo/semantics/openAccess
Creative Commons Attribution Non Commercial 4.0 International
https://creativecommons.org/licenses/by-nc/4.0/legalcode
dMRI, Tractography, White Matter, Human Connectome Project, Diffusion MRI, Connectomics, Segmentation
High quality white matter reference tracts
info:eu-repo/semantics/other