Fiber Tracking
use –action=trk to initiate fiber tracking in DSI Studio.
The fiber tracking is a deterministic tracking approach doeumented in Yeh, Fang-Cheng, et al. “Deterministic diffusion fiber tracking improved by quantitative anisotropy.” PloS one 8.11 (2013): e80713.
The pipeline also includes computations for connectivity matrices and tractometry.
For mapping individual bundle, please use (automatic fiber tracking)
Examples
Track the left arcuate fasciculus a fib files and save them in tractography
dsi_studio --action=trk --source=*.fib.gz --track_id=Arcuate_Fasciculus_L --output=*.AF.tt.gz
Perform fiber tracking with all fiber tracking parameters assigned by a parameter ID
dsi_studio --action=trk --source=subject001.fib.gz --parameter_id=c9A99193Fba3F2EFF013Fcb2041b96438813dcb
Use the left and right precentral region from FreeSurferDKT atlas as the ROIs. Dilate them twice and smooth them for fiber tracking
dsi_studio --action=trk --source=subject001.fib.gz --roi=FreeSurferDKT_Cortical:left_precentral,dilate,dilate,smoothing --roi2=FreeSurferDKT_Cortical:right_precentral,dilate,dilate,smoothing
Fiber tracking using two MNI-space ROIs (include MNI in the file name) and subject-space whole-brain seeding (from wholeBrain.nii)
dsi_studio --action=trk --source=subject001.fib.gz --seed=wholeBrain.nii --roi=mni_roi1.nii --roi2=mni_roi2.nii
Perform fiber tracking and output connectivity matrix using HCP-MMP and AAL3 parcellation, respectively. Specify –output=no_file to skip saving the large tractography file
dsi_studio --action=trk --source=subject001.fib.gz --fiber_count=1000000 --output=no_file --connectivity=HCP-MMP,FreeSurferDKT_Cortical --connectivity_value=count,qa,trk
Perform fiber tracking and export track-specific statistics and TDI
dsi_studio --action=trk --source=subject001.fib.gz --output=tracks.tt.gz --export=stat,tdi
Perform fiber tracking and insert other metrics (DKI, FW_FA…etc) and export track-specific statistics
dsi_studio --action=trk --source=subject001.fib.gz --other_slices=./other/*.nii.gz --output=tracks.tt.gz --export=stat
Perform differential tractography by comparing MNI-space FA map (include MNI in the file name) and subject’s fa and tracking the decreased FA at 10%
dsi_studio --action=trk --source=subject001.fib.gz --other_slices=template_qa.nii.gz --dt_threshold_index=template_qa-qa --dt_threshold=0.2 --seed_count=1000000 --min_length=30 --output=tracks.tt.gz
(Windows System) Search for all fib files under the current directory and perform fiber tracking to get the FreeSurferDKT connectivity matrix
path=C:\dsi_studio_64
for /f "delims=" %%x in ('dir *.fib.gz /b /d /s') do (
call dsi_studio.exe --action=trk --source="%%x" --seed_count=1000000 --thread_count=8 --output=no_file --connectivity=FreeSurferDKT > "%%x.log.txt"
)
Core Functions
| Parameters | Default | Description |
|---|---|---|
| source | specify the fib file | |
| thread_count | hardware max | specify the thread count. |
| track_id | (not assigned) | specify which ID of the track to track using automatic fiber tracking. The complete list of ID is found in ICBM152.tt.gz.txt in the DSI Studio package under the \atlas\ICBM152 folder (On Mac, right click to show content). |
Conventional Tracking
Specify the tracking parameters below or replace them by using a single
--parameter_id, which can be found at the method text under Step T3d after running fiber tracking.
| Parameters | Default | Description |
|---|---|---|
| fiber_count or seed_count | 100000 |
specify the number of fibers to be generated. If seed number is preferred, use seed_count instead. If DSI Studio cannot find a track connecting the ROI, then the program may run forever. To avoid this problem, you may assign fiber_count and seed_count at the same time so that DSI Studio can terminate if the seed count reaches a large number. |
| fa_threshold | 0 |
which means it is randomized): threshold for fiber tracking. In QBI, DSI, and GQI, “fa_threshold” will be applied to the QA threshold. To use other index as the threshold, add “threshold_index=[name of the index]” (e.g. “–threshold_index=nqa –fa_threshold=0.01” sets a threshold of 0.01 on nqa for tract termination). If fa_threshold is not assigned, then DSI Studio will select a random value between 0.5Otsu and 0.7 Otsu threshold using a uniform distribution. |
| turning_angle | 0 (randomized) |
This threshold (in degrees) serves as a termination criterion. If two consecutive moving directions have a crossing angle above this threshold, the tracking will be terminated. The default 0 will be a random selection of a value from 15 degrees to 90 degrees. |
| step_size | 0 (randomized) |
Step size defines the moving distance in each tracking iteration. This unit is in millimeter-scale. The default value 0 will be a random selection of the step size from 0.5 to 1.5 voxel distance. |
| min_length | 30 (mm) |
Length constraint that filters out the tracks that are either too short |
| max_length | 300 (mm) |
Length constraint that filters out the tracks that are too long |
The following settings are also included in --parameter_id but usually the default values are used.
| Parameters | Default | Description |
|---|---|---|
| method | 0 |
tracking methods 0:streamline , 1:rk4 |
| otsu_threshold | 0.6 |
The default Otsu’s threshold can be adjusted to any ratio. |
| initial_dir | 0 |
initial propagation direction 0:primary fiber, 1:random, 2:all fiber orientations. The Primary approach always starts the tracking from the most prominent fiber in a voxel. The advantage is the stableness and consistency of the results. A way to compensate for this drawback is using whole-brain seeding to explore all possible connections. The Random approach starts the tracking a randomly selected fiber orientation, so the results have random factors. The advantage is that it can explore all possibilities, but the drawback is that the results may not be reproduced exactly. The tracking results are also sensitive to noisy fibers because the false fiber orientation can be selected as the starting direction. The All approach starts a track for each fiber orientation resolved in a voxel. It aims to explore all possible connections and there are no stochastic factors that may hurt the reproducibility. The drawback is that it is sensitive to noisy fibers because all resolved fiber orientations will be used as the starting directions. |
| smoothing | 0 (off) |
Smoothing serves like a “momentum”. For example, if smoothing is 0, the propagation direction is independent of the previous incoming direction. If the smoothing is 0.5, each moving direction remains 50% of the “momentum”, which is the previous propagation vector. This function makes the tracks appear smoother. In implementation detail, there is a weighting sum on every two consecutive moving directions. For smoothing value 0.2, each subsequent direction has 0.2 weightings contributed from the previous moving direction and 0.8 contributed from the income direction. To disable smoothing set its value to 0. Assign 1.0 to do a random selection of the value from 0% to 95%. |
| tip_iteration | 0 |
specify pruning iterations. If –track_id or –dt_threshold_index is specified, the default value is 16 |
Please note that parameter ID does not override ROI settings, dt_threshold_index, or threshold_index settings. You may still need to assign ROI/ROA…etc. in the command line.
ROI Settings
| Parameters | Description |
|---|---|
| seed | specify the seeding file. Supported file format includes text files and nifti files. |
| roi roi2 roi3 roi4 roi5 | specify the roi files. Supported file format includes text files and nifti files. |
| roa roa2 roa3 roa4 roa5 | specify the roa files. |
| end end2 | specify the endpoint files. |
| ter ter2 ter3 ter4 ter5 | specify the terminative region. |
| t1t2 | specify t1w or t2w images as the ROI reference image |
Multiple input is supported. You may use atlas regions as the roi, roa, end, or ter regions. Specify the atlas and region name (case sensitive!) as follows:
--roi=FreeSurferDKT:left_precentral
An ROI (and also other region types) can be modified by the following action code: “smoothing”, “erosion”, “dilation”, “defragment”, “negate”, “flipx”, “flipy”, “flipz”, “shiftx” (shift the roi by 1 in x direction), “shiftnx” (shift the roi by -1 in x direction), “shifty”, “shiftny”, “shiftz”, “shiftnz”.
--roi=region.nii.gz,dilation,dilation,smoothing #This dilates the region in the region.nii.gz file twice and smooth itYou can assign the region value to be loaded
--roi=multiplre_region_nifti.gz:1 # only loads regions with value=1
You can assign an MNI space NIFTI file. Just make sure to have MNI in the file name
--roi=mni_broca1.nii.gz
Differential tracking settings
| Parameters | Description |
|---|---|
| other_slices | [option 1] specify the NIFTI file or to be inserted for differential tractography (e.g., –other_slices=pre.nii.gz,post.nii.gz) [option 2] specify a connectometry database with built-in demographics and specify subject’s demographics using –subject_demo (e.g., –other_slices=study.db.fib.gz –subject_demo=62,1) |
| dt_threshold | assign percentage threshold for differential tractography. assign –dt_threshold=0.1 to detect more than 10% change. |
| dt_threshold_index | specify the metrics for differential tracking (e.g., –dt_threshold_index=post-pre) |
| dt_threshold_type | specify the calculation of the differences. 0: (m1-m2)/m1 1:(m1-m2) |
To load slices as “MNI images”, please include “mni” in the file name
--other_slices=mni_qa.nii.gz
Tract-related post-processing and analysis
| Parameters | Description |
|---|---|
| delete_repeat | assign the distance for removing repeat tracks (e.g. –delete_repeat=1 removes repeat tracks with distance smaller than 1 mm) |
| output | specify the output directory or output file name (tt.gz, trk.gz, or nii.gz ). Specify no_file to disable tractography output. |
| end_point | specify file name for output endpoint coordinates (.txt or .mat) |
| template_track | specify file name for output tracts in the tempalte space |
| export | export additional information related to the fiber tracts <p> use “–export=tdi” to generate track density image in the diffusion space. <p> use “–export=tdi2” to generate track density image in the subvoxel diffusion space. <p> use “–export=tdi_color” or “–export=tdi2_color” to generate track color density image. <p> use “–export=stat” to export tracts statistics like along tract mean fa, adc, or morphology index such as volume, length, … etc. <p> <p> To export TDI endpoints, use tdi_end or tdi2_end. <p> <p> use “–export=report:dti_fa:0:1” to export the tract reports on “fa” values with a profile style at x-direction “0” and a bandwidth of “1” <p> the profile style can be the following: <p> <p> 0 x-direction <p> 1 y-direction <p> 2 z-direction <p> 3 along tracts <p> 4 mean of each tract <p> <p> You can export multiple outputs separated by “,”. For example, <p> –export=stat,tdi,tdi2 exports tract statistics, tract density images (TDI), subvoxel TDI, along tract qa values, and along tract gfa values. |
| ref | output track coordinate based on a reference image (e.g. –ref=T1w.nii.gz |
Connectivity analysis
| Parameters | Default | Description |
|---|---|---|
| connectivity | output connectivity matrix using ROIs or atlas as the matrix entry. | |
| connectivity_threshold | 0.001 |
specify the threshold ( in relative ratio to the max value) for calculating binarized graph measures and connectivity values. This means if the maximum connectivity count is 1000 tracks in the connectivity matrix, then at least 1000 x 0.001 = 1 track is needed to pass the threshold. Otherwise, the values will be set to zero. |
| connectivity_type | pass |
specify whether to use “pass” or “end” to count the tracks. |
| connectivity_value | count |
specify the way to calculate the matrix value. count outputs the number of tracks passing/ending in the regions. ncount outputs the number of tracks normalized by the median length. mean_length outputs the mean length of the tracks. trk outputs a trk file each connectivity matrix entry. dti_fa outputs mean FA qa outputs mean QA. You can output any metrics or even metrics added using –other_slice. You can output multiple results by separating the parameters with “,”. (e.g. --connectivity_value=count,ncount,trk) |
| connectivity_output | matrix,connectogram,measure |
specify whether to output connectivity matrix (matrix), connectogram, or network measures (measure) |
If you would like to use the built-in atlas, please specify the atlas name.
--connectivity=FreeSurferDKT_cortical,HCP-MMPIf you would like to use subject space ROIs, specify the file name
--connectivity=/subject01/segmentation_in_dwi_space.nii.gzIf you would like to use your customized MNI space atlas, specify the file name (must have
mniin the file name)--connectivity=mni_space_parcellations.nii.gzIf the ROIs is segmented based on T1W, you will need to specify the original t1w file using –t1t2
--t1t2=subject_t1w.nii.gz --connectivity=parcellation_based_on_subject_t1w.nii.gzPlease note that if the number of the connecting tracks is not enough (determined by connectivity_threshold), the connectivity value will not be calculated. This strategy is used to avoid the inclusion of accidental connections due to false tracks.
Clustering settings
| Parameters | Default | Description |
|---|---|---|
| cluster | optional | run track clustering after fiber tracking. 4 parameters separated by comma are needed: –cluster=METHOD_ID,CLUSTER_COUNT,RESOLUTION,OUTPUT FILENAME METHOD_ID: 0=single-linkage clustering 1=k means 2=EM CLUSTER_COUNT: The total number of clusters. In k-means or EM, this is the total number of clusters assigned. In single-linkage, it is the maximum number of clusters allowed to avoid over-segmentation (the remaining small clusters will be grouped in the last clusters). RESOLUTION: only used in single-linkage clustering. The value is the mini meter resolution for merging the clusters. OUTPUT FILENAME: the file name for output the cluster label (should be a text file). The file name should not contain a space. |
--cluster=0,500,2,label.txt #run a single-linkage cluster with 500 maximum cluster count and 2-mm resolution, saved as label.txt