Hi Ruth
They differ in the quantities they estimate and the way they perform inference. Bedpostx estimates a simplified version of the fibre ODF, what you really want for tractography, in a parametric way. Qboot estimates currently only versions of the diffusion ODF, a blurred version of the fODF, in a non-parametric way. Therefore bedpostX gives you greater sensitivity in crossings when the model assumptions hold. These assumptions are good enough for most typical single and multi-shell HARDI schemes and I would suggest bedpostX -model 2 for a multishell acquisition.
Cheers
Stam
On 2 Jul 2014, at 07:07, Ruth <[log in to unmask]> wrote:
> Hi,
> I'm in the early stages of collecting some multi-shell diffusion data and am assessing analysis options. What are the tradeoffs between qboot (ODFs) and bedpost (PDFs)? Can you point me to a good overview article for a non-mathematician?
|