Revert to read bval from PVM_DwEffBval and bvec from PVM_DwGradVec

[jeremie-fouquet]

Fixes #90 and #95.

Changes proposed in this pull request:

• In _get_bdata, revert to reading bval from PVM_DwEffBval and bvec from PVM_DwGradVec. Both of these parameters should include the b0 scans correctly ordered, which solves Ordering of BVAL/BVEC in DTI data #90. PVM_DwEffBval also contains the effective b-values, which include the effect of all applied gradients; it solves Outputted diffusion bval are not the effective b-values anymore #95.

To consider: according to Bruker documentation for PV6.0.1 and PV360, PVM_DwGradVec is a parameter "specifying the diffusion gradient amplitudes in the x,y,z coordinate system". This coordinate system (the scanner one) does not correspond to the FSL convention, which brkraw seems to be using. I'll open another issue on that topic.

@BrkRaw/Bruker

[dvm-shlee]

@jeremie-fouquet Do effective bvals and bvacs from this code match the order to the volume generated if the scan is applied repeatedly? If you confirm this, I will apply this change!

[jeremie-fouquet]

Hi @dvm-shlee.

For all the data sets I tested, effective bvals from PVM_DwEffBval and bvecs from PVM_DwGradVec were correctly ordered with respect to the acquisition volumes. This is true even for B0 volumes (no need to do additional corrections to the bvals and bvecs for B0 acquisitions).