dsi2 documentation

This software is designed to enable the analysis of diffusion-weighted MRI-based tractography datasets. DSI2 stands for D iscrete S patial I ndexing of D iffusion S pectrum I mages. Image reconstruction, tractography and parcellation must be done outside of this software, but indexing, searching and analysis are all implemented here. Specifically, this package provides the tools necessary to perform Local Termination Pattern Analysis (LTPA, [1] )

The idea is that

1. A set of coordinates in the brain can be used to query a database.
2. The query returns a set of streamlines that that pass through the set of coordinates.
3. The streamlines are divided into groups based on
   1. Their termination regions
   2. Clustering based on their shape
4. The groups of streamlines are compared across individuals

Take, for example, the corticospinal tract. If we select a sphere of coordinates in the internal capsule, we should see tracks following the pathway of the corticospinal tract.

We can then group these streamlines based on the regions in which they terminate. Using the Lausanne2008 scale 33 atlas, we get the following distribution of streamline counts:

Getting Started

Installing DSI2

Preparing data

Use DSI Studio and connectome_mapper to prepare data

Organizing your local datasource (Hard way)

An overview of possible analyses and pitfalls to beware of.

• Installing DSI2
  • Dependencies
    • Even if using Canopy, you will need to install the following packages:
    • Next, you will be using Canopy’s “Package Manager” in the welcome screen as shown below:
    • Open your terminal to download the DSI2 source code:
  • Setting up your environment
    • Starting MongoDB
    • Verifying your installation
• DSI2 Download Details
  • FreeSurfer
    • XQuartz
  • Easy Lausanne
  • Installing DSI Studio
• Preparing data
  • dsistudio
  • connectomemapper
  • mongodb
• Preparing Data with Freesurfer
• DSI2 Reconstruction
  • DSI Studio
    • Step 1
    • Step 2
    • Reconstruction
    • Step 3
    • Step 4: Free Surfer
    • Step 5: Easy Lausanne
    • Step 6: Checking Registration
    • CORONAL:
    • SAGITTAL:
    • HORIZONTAL:
    • STEP 7: Dilate your images by 1 voxel
    • STEP 8: Comparing your files
• Organizing your local datasource (Hard way)
  • Basic information
  • Streamline Labels
    • Creating streamline labels
• Importing your data
• Accessing and analyzing your data
  • Searching your local datasource
  • Searching your MongoDB Datasource
  • Aggregating streamlines based on termination regions
  • Running a whole-brain LTPA
  • References
• Creating a streamline aggregator
  • Defining the Aggregator class
    • Defining algorithm parameters as traits
    • Specifying the GUI editors for your aggregator
    • Overriding the .aggregate() method
  • Exploring your aggregator in realtime
• Browsing the Streamline Database
  • Red Text: Data Source and Aggregator
    • Select your Data Source in the top left of the screen
    • Change Aggregation Algorithm to Region Labels
  • Yellow Box: Specifying data source properties
  • Blue Box: Matching datasets
• The Sphere Browser: Querying and visualizing streamlines
  • Selecting Coordinates: Using a sphere
  • Selecting Coordinates: From an ROI
  • Visualizing Streamlines
  • Changing the appearance of streamlines

References

[1]	Cieslak, M., & Grafton, S.T. Local termination pattern analysis: a tool for comparing white matter morphology. Brain Imaging Behav, DOI 10.1007/s11682-013-9254-z (2013).

Indices and tables

• Module Index
• Search Page