Hi - the distance threshold works on each half of the streamline separately.
On 5 Jun 2012, at 07:09, Colin Reveley wrote:
> If a single track is seeded from a single voxel, in fact two tracks are seeded that proceed in exactly opposite directions from the seedpoint.
> If I specify that tracks whose length is less than 5mm are to be discarded does this refer to the combined length? how does this interact with mask intersections?
> If one segment of the sample proceeds towards a rostral mask, and the other to a caudal mask, will the entire sample be discarded if it hits the caudal mask when the combined sample length is <5mm? Or will the other segment be allowed to proceed towards the rostral mask?
> if the other segment is allowed to proceed, then the total sample length may then exceed 5mm. And so, does the intersection of the sample with the caudal (<5mm from seed) mask count, by virtue of the total sample length being greater than 5mm including the rostral extent?
> How exactly does this work, in terms of specifying a seed to classification mask set, when one wants the distance travelled from seed to mask to be at least 5mm, for the seed to get a point for hitting the mask?
> I would say in principle the two segments are the same sample. Yet this seems to pose problems if you want to specify that a track from a seed point to a mask must have a certain length. It would be good at least to know what the parameter specifies precisely, in seed to target terms rather than just sample length terms.
University of Oxford, FMRIB Centre
JR Hospital, Headington, OX3 9DU, UK
(+44)1865-222466 (fax 717)