Understand existing Slicer DMRI ecosystem

[ebrahimebrahim]

• Understand well what Slicer and SlicerDMRI already can do.
• Think about how that affects the design of our Slicer modules

---

I did a survey myself in July 2023, here is my summary.

Slicer has a DWI node which can see b-value info and let you scroll through the components.

It has a DTI node which can be visualized in many different ways: directionally encoded RGB, derived scalars such as FA, ellipsoid or principal eigenvector glyphs, and more.

NRRD is Slicer's native way to read and write both DWI and DTI in Slicer, but DICOM is also supported in certain ways by certain modules.

There is some very simple QC tooling in Slicer. The more serious QC tooling from UNC NIRAL seems to have fallen out of support.

There is a concept of a fiber bundle node in Slicer, and these can be saved and loaded as vtk files and they can be generated from DWI or DTI using the tractography tools.

All tractography is determinstic (though the seeding can be randomized). It is based on picking initial seed regions and then generating streamlines from there. The streamlines can be generated using DTI, or there is the UKF tractography module that can use a 2-tensor model, a NODDI-based model with up to two orientations, and that can also include a separate free water compartment.

[ebrahimebrahim]

We are not diving into slicer ecosystem integration yet