Skip to content

aledelmo/TractographyMetrics

Repository files navigation

Tractography Analysis

Neural connections geometrical and functional analysis

tractography

Compute multiple statistics about fiber bundles. Descriptive statistics and quantitative information about image information and geometrical characteristics.

  • Bundle properties
    • Number of fibers: number of streamlines composing the tractogram.
    • Number of points: number of points composing a streamline
  • Geometrical properties
    • Lengths: length of the fiber (in mm). Computed by adding up the lengths of all segments. Feature: EDT
    • Shortest path: length of the straight segment connecting fibers endpoints
    • Turning angle: turning angle projected (Winding). Cumulative signed angle between each line segment and the previous one, expressed in degrees.
  • Connectivity properties
    • Seeds position: position of the starting point (seed) of a streamline
    • Termination position: position of the endpoint (termination condition respected) of a streamline
    • Mid-point position: position of the central point of a streamline
  • Diffusion properties
    • b-zero intensity: mapping of every point of the streamline on the baseline image. The exact mapped value is computed with a weighted interpolation of the b-zeros of the 12-connected voxel space around the point.
    • FA value: Mapping of every point of the streamline on the FA map.
    • MD value: Mapping of every point of the streamline on the MD map.

Installation and Usage.

Installation:

$ git clone https://github.com/aledelmo/TractographyMetrics
$ cd TractographyMetrics
$ pip install -r requirements.txt
$ python -m tractography_metrics.py <tractogram_filepath> <output_txt_file>

To enrich analysis with diffusion information use the following optional keywords:

short flag long flag Action
-bzero <filepath> --b_zero <filepath> Compute stats on the bzero volume
-fa <metric_filepath> --Fractional_Anisotropy <metric_filepath> Compute stats on the FA metric volume
-md <metric_filepath> --metric_filepath Compute stats on the MD metric volume

Save the output in a different format with the optional flags:

short flag long flag
-csv --save_csv
-xlsx --save_xlsx

Add a header:

short flag long flag Action
-hd <text> --header <text> Add any string of text to the stats

When working on huge amounts of fibers you might experience computational issues. Try to downsample the tractogram with:

short flag long flag Action
-r <resampling_percentage> --resample <resampling_percentage> Specify the downsample percentage of the tractogram fibers (value between 0 and 100)

Contacts

For any inquiries please contact: Alessandro Delmonte @ alessandro.delmonte@institutimagine.org

License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details

About

Neural connections geometrical and functional analysis

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages