Notes from informal meeting about diffusion
On 10 May 2007
Acquisition issues
Need for optimization for fibre tracking
Issues are:
- Number of averages vs number of directions
- Isotropic vs defined diffusion angles - relating to anatomy
- Cardiac gating; expected differences; whether this can be easily applied using Siemens setup. Is currently working for Oxford with Siemens patches - these may not be available to us
CBU acquisitions
- Default is 63 angles, 2 averages, 14 minutes
- Alternative 30 directions 3 or 4 averages
Need to investigate which is preferable for fibre tracking
WBIC
For fibre tracking:
- 8 minutes, 2mm isotropic, 63 directions - ? one average (as it were)
For traumatic brain injury patients (Virginia Newcombe):
- 12 direction, 5 B values (allowing more accurate estimation of diffusion coefficent)
- Physicists involved are Guy Williams + Adrian Carpenter
Analysis
CBU currently using FSL - see DiffusionFdtNotes
BrainVisa set up, but display not working due to OpenGL problems. Needs local (desktop) install
WBIC using dtiquery - http://graphics.stanford.edu/projects/dti/dti-query/ good display routines, fibre tracking - algorithm? Have also used FSL - particularly for fibre track cross subject analysis using TBSS program (see below).
Open questions
How to do quantitative cross-subject analysis or fibre track anatomy in relationship to neuropsychology? Investigation of TBSS? Using FSL fibre tracking for statistical analysis across subjects of connectivity between areas
Plan
Hope to set up more knowledge sharing between WBIC and CBU / CSL. Maybe Wiki pages for WBIC on diffusion experience and software.
First step at understanding cross-subject analyses? Suggest journal club on TBSS main paper.
S.M. Smith, H. Johansen-Berg, M. Jenkinson, D. Rueckert, T.E. Nichols, K.L. Miller, M.D. Robson, D.K. Jones, J.C. Klein, A.J. Bartsch, and T.E.J. Behrens. Acquisition and voxelwise analysis of multi-subject diffusion data with Tract-Based Spatial Statistics. Nature Protocols, 2(3):499-503, 2007. Technical report: http://www.fmrib.ox.ac.uk/analysis/research/tbss/tr05ss1.pdf