[Mrtrix-discussion] seed point generation

[Donald Tournier]

Hi Maarten,

When seeding from a seed image, each seed point is generated at random
from within the volume of interest, which is defined as the set of
points whose nearest neighbouring voxel has intensity greater than
0.5. Another way to think about it is that the software first selects
a voxel at random from the set of non-zero mask voxels, then generates
a random location within the cuboid volume corresponding to that
voxel. For example, if the voxel at integer voxel coordinates [ x y z
] is selected, the seed point will be generated at random from the
interval [ x-0.5 => x+0.5 ; y-0.5 => y+0.5; z-0.5 => z+0.5 ].

In more detail, what happens is: the code generates a random
floating-point location within the volume of the mask image with
uniform probability, converts that location to pixel coordinates,
rounds these coordinates to integer voxel coordinates, and if the
intensity of that voxel is greater than 0.5, the seed point is
accepted. Otherwise the process is repeated until a valid seed point
has been generated.

Hope that clarifies the process.
Cheers,

Donald.


On Tue, Nov 3, 2009 at 1:22 AM, Maarten Vaessen <mjvaessen at yahoo.com> wrote:
> Hello,
>
>
>
> I see from de documentation that when specifying a seed point for tracking,
> the “tracks are seeded at random from a spherical ROI position at [ -5.3 17
> -30.7 ] with a 3 mm radius”. This makes sense of course, but how does this
> work when specifying a seed mask from, for instance, an analyze image. Say,
> the seed image has 100 non-zero voxels and I specify streamtrack with
> –maxnum 10000, how are the exact seed points generated within the seed
> voxels?
>
>
>
> Kind regards Maarten
>
> _______________________________________________
> Mrtrix-discussion mailing list
> Mrtrix-discussion at www.nitrc.org
> http://www.nitrc.org/mailman/listinfo/mrtrix-discussion
>
>



--
Jacques-Donald Tournier (PhD)
Brain Research Institute, Melbourne, Australia
Tel: +61 (0)3 9496 4078



-- 
Jacques-Donald Tournier (PhD)
Brain Research Institute, Melbourne, Australia
Tel: +61 (0)3 9496 4078