[Camino-users] track - slight change to anisotropy thresholds in r1061

Philip A Cook cookpa at mail.med.upenn.edu
Fri Jun 15 07:26:58 PDT 2012


Hi,

Previously it was impossible to apply an anisotropy threshold in voxels containing multiple compartments. This has been changed in r1061. 

Single tensor tractography is unchanged.

For multi-tensor data, the anisotropy threshold is applied to the average anisotropy, so tracking terminates if the mean FA is below the threshold.

For other input types to track, when an image is passed with -anisfile, the value in the file is applied to all voxels. Previously, the anisotropy threshold was not used if the voxel contained multiple fiber orientations. This was by design since DT FA is usually underestimated in such voxels, but it made it impossible for users to filter out voxels where a higher order model was erroneously fitted to the data.


Phil


More information about the Camino-users mailing list