How to adapt coherency to 3D?

[chourroutm]

Hi,
Thank you for the great tool! I came across the PDF document with the methods. How does the coherency computation transfer to 3D? Is it equal to what the MRI community refers to as fractional anisotropy (FA)?

[dasv74]

Thank you for your message. This is an interesting topic that could also be posted on the image.sc forum.

We now also have a 3D version of Orientation in Python, which you can find here:
https://epfl-center-for-imaging.gitlab.io/orientationpy/introduction.html

In this version, we have defined two types of coherency (or coherence) to capture different features in the data volume:

• Coherency of fiber or tube: This is high when λ_max is much larger compared to the average of the other two eigenvalues.
• Coherency of membrane or plane: This is high when λ_min is much smaller compared to the average of the other two eigenvalues.