>With VB15, mriconvert version 2.0 Revision 190 extracts
>exactly what is in the DiffusionGradientDirection field
yes. But this entry is estimated from the b-matrix and can - due to a bug - in some instances can yield an inverse solution. Which does not matter for FA etc but it does for probabilistic tracking and the like. This has been fixed from VB17 on. Prior to thatit is safer to use no angulation + gradients in XYZ.
PS: It does, of course, NOT help to load VB15 data into a VB17 and to safe the data again from VB17.
Von: FSL - FMRIB's Software Library [[log in to unmask]] im Auftrag von Josef Ling [[log in to unmask]]
Gesendet: Donnerstag, 24. Juni 2010 02:27
An: [log in to unmask]
Betreff: Re: [FSL] AW: [FSL] AW: [FSL] DTI bvecs - dcm2nii vs. mriconvert vs. dtistudio vs. siemens VB15 official tables
i'd like to get this thread going again and check an assumption.
i'm reading Siemens data from a VB15 version triotim. i have been assuming that the true gradient vector is stored in the dicom header under the DiffusionGradientDirection field. i've been using that to check the accuracy of the output from mriconvert and dcm2nii. Does anyone know if that is a safe assumption?
With VB15, mriconvert version 2.0 Revision 190 extracts exactly what is in the DiffusionGradientDirection field.
- josef ling