Thanks. Just to make sure, how is the fslstats script written to get the value M from a fdt_path.nii.gz image file? I imagine you would need to figure out all the voxels in the image and sum up their individual intensities.
My best,
aj
--- On Mon, 7/18/11, Stamatios Sotiropoulos <[log in to unmask]> wrote:
> From: Stamatios Sotiropoulos <[log in to unmask]>
> Subject: Re: [FSL] Probabilistic tractography
> To: [log in to unmask]
> Date: Monday, July 18, 2011, 5:56 AM
> The intensity in each voxel of the
> image represents this number.
>
> Cheers,
> Stam
>
>
> On 18 Jul 2011, at 13:04, Adil Javed wrote:
>
> > Thanks so much. Since fdt_paths.nii.gz is an
> image file, how do you get the value M?
> >
> > many thanks,
> > AJ
> >
> > --- On Mon, 7/18/11, Stamatios Sotiropoulos <[log in to unmask]>
> wrote:
> >
> >> From: Stamatios Sotiropoulos <[log in to unmask]>
> >> Subject: Re: [FSL] Probabilistic tractography
> >> To: [log in to unmask]
> >> Date: Monday, July 18, 2011, 3:04 AM
> >> Hi Aj,
> >>
> >> The "connectivity value" in probabilistic
> tractography from
> >> a seed A to a voxel B is defined as the ratio M/N,
> where M
> >> is the number of curves that start from A and go
> through B
> >> and N the total number of curves starting from A.
> This ratio
> >> gives an estimate of the probability of the path
> of least
> >> hindrance to diffusion.
> >>
> >> The waytotal file gives you N, the total number of
> *valid*
> >> curves that have started from A (after excluding
> curves that
> >> do not meet exclusion/inclusion criteria). M for
> each voxel
> >> in the image is described in the fdt_paths file.
> Dividing
> >> the two will give path probabilities.
> >>
> >> Hope this is clear,
> >> Stam
> >>
> >>
> >> On 18 Jul 2011, at 09:04, Adil Javed wrote:
> >>
> >>> Hi,
> >>> I have generated some data using tractography.
>
> >> The waytotal text file shows a number. Is it
> more
> >> accurate to report that number as connectivity
> value or an
> >> estimate of number of fibers per seed/mask on a
> plot?
> >>>
> >>> many thanks,
> >>>
> >>> aj
> >>>
> >>
> >
>
|