[Insight-users] threshold at output of CollidingFrontsImageFilter

[Luca Antiga]

Hi Bryn,
  in vmtk I keep the threshold fixed to -10 *  
collidingFrontsNegativeEpsilon, but if you come up with a smarter  
threshold, just let me know.
Keep in mind that colliding fronts was created to initialize level  
sets or deformable models. It's handy because it selectively  
initializes branches without side branches, but it may not be very  
accurate in catching the location of the boundary of the vessel (this  
very much depends on the image quality - if the signal in the vessel  
is high, it can do a pretty good job).

Last, you might be interested in knowing that if you add

fastMarchingFilter1->SetTargetReachedModeToOneTarget();

and

fastMarchingFilter2->SetTargetReachedModeToOneTarget();

in itkCollidingFrontsImageFilter.txx the filter will run sensibly  
faster.

It's in my todo list to add this as an extra option to the filter in  
ITK.

Best regards

Luca




On Feb 13, 2009, at 4:18 PM, Bryn Lloyd wrote:

> Hi Luca and Dan,
>
>
> Is there any heuristic, which can help to choose a threshold on the  
> output of the CollidingFrontsImageFilter?
>
> I have following pipeline:
>
> MRA image -> reader -> sigmoid -> colliding fronts -> threshold ->  
> writer.
>
> Unfortunately, the (subjective) optimal iso-level/threshold seems to  
> vary quite a bit from case to case.
>
>
> Thanks
> Bryn
>

--
Luca Antiga, PhD
Head, Medical Imaging Unit,
Biomedical Engineering Department,
Mario Negri Institute.
mail: Villa Camozzi, 24020, Ranica (BG), Italy
phone: +39 035 4535-381
email: antiga at marionegri.it
web: http://villacamozzi.marionegri.it/~luca

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20090213/7d551ab2/attachment-0001.htm>