[Insight-users] Centerline

[Dan Mueller]

Hi Simon,

I typically pad the image with lower=[5,5,5] and upper=[5,5,5], though
I think a padding of 2 is theoretically sufficient.

HTH

Cheers, Dan

On 30 April 2012 18:37, Simon ESNEAULT <simon.esneault at gmail.com> wrote:
> Hi Dan and all,
>
> We use here your nice contribution for the centerline extraction, and we
> also had some trouble with centerlines near the border, with a crash and
> that kind of backtrace :
>
> 1 - 0x00000001001a0d28: itk::InterpolateImageFunction<itk::Image<float, 3u>,
> double>::Evaluate(itk::Point<double, 3u> const ) const in
> libThEndosize.dylib
> 2 - 0x0000000100194eb4:
> itk::PhysicalCentralDifferenceImageFunction<itk::Image<float, 3u>,
> double>::Evaluate(itk::Point<double, 3u> const ) const in
> libThEndosize.dylib
> 3 - 0x00000001001951c4: itk::SingleImageCostFunction<itk::Image<float, 3u>
>>::GetDerivative(itk::Array<double> const , itk::Array<double> ) const in
> libThEndosize.dylib
> 4 - 0x0000000100620ae1:
> itk::RegularStepGradientDescentBaseOptimizer::ResumeOptimization() in
> libThItk.dylib
> 5 - 0x0000000100620f62:
> itk::RegularStepGradientDescentBaseOptimizer::StartOptimization() in
> libThItk.dylib
> 6 - 0x00000001001ac692: itk::ArrivalFunctionToPathFilter<itk::Image<float,
> 3u>, itk::PolyLineParametricPath<3u> >::GenerateData() in
> libThEndosize.dylib
> 7 - 0x00000001001b0651: itk::SpeedFunctionToPathFilter<itk::Image<float,
> 3u>, itk::PolyLineParametricPath<3u> >::GenerateData() in
> libThEndosize.dylib
>
> Our assumption was also that in some case when the centerline was near the
> border the gradient descent optimizer ask for a point outside of the image,
> leading to that crash, but we disdn't use the "padding" hack. So my question
> is : how much should be the padding to be sure that a crash like that won't
> occurs ?
>
> Thanks again for your contribution.
>
> -Simon
>
>
> On Sun, Apr 29, 2012 at 00:28, Dan Mueller <dan.muel at gmail.com> wrote:
>>
>> Hi Charu,
>>
>> (I was just re-reading my email, I don't think I was clear enough.)
>>
>> I used Fast Marching to extract the paths in your 3d dataset. Then I
>> projected both the dataset and the paths onto a 2d image, which was
>> attached to my previous email. In short: the method works for 3d data.
>>
>> Cheers, Dan
>>
>> On 29 April 2012 07:54, Dan Mueller <dan.muel at gmail.com> wrote:
>> > Hi Charu,
>> >
>> > The image I attached to my previous email was a MIP of your 3d
>> > dataset. This method works with both 2d or 3d data.
>> >
>> > Cheers, Dan
>> >
>> > On 29 April 2012 05:17, chans at cs.uh.edu <chans at cs.uh.edu> wrote:
>> >> Hi Dan,
>> >>
>> >> Thank You for your reply. I looked at Fast Marching minimization. It
>> >> works
>> >> for 2d images. What, I have here is 3d data.
>> >>
>> >> Thank You
>> >> Charu
>> >>
>> >> --
>> >> View this message in context:
>> >> http://itk-insight-users.2283740.n2.nabble.com/Centerline-tp7507501p7509617.html
>> >> Sent from the ITK Insight Users mailing list archive at Nabble.com.
>> >> _____________________________________
>> >> Powered by www.kitware.com
>> >>
>> >> Visit other Kitware open-source projects at
>> >> http://www.kitware.com/opensource/opensource.html
>> >>
>> >> Kitware offers ITK Training Courses, for more information visit:
>> >> http://www.kitware.com/products/protraining.php
>> >>
>> >> Please keep messages on-topic and check the ITK FAQ at:
>> >> http://www.itk.org/Wiki/ITK_FAQ
>> >>
>> >> Follow this link to subscribe/unsubscribe:
>> >> http://www.itk.org/mailman/listinfo/insight-users
>> _____________________________________
>> Powered by www.kitware.com
>>
>> Visit other Kitware open-source projects at
>> http://www.kitware.com/opensource/opensource.html
>>
>> Kitware offers ITK Training Courses, for more information visit:
>> http://www.kitware.com/products/protraining.php
>>
>> Please keep messages on-topic and check the ITK FAQ at:
>> http://www.itk.org/Wiki/ITK_FAQ
>>
>> Follow this link to subscribe/unsubscribe:
>> http://www.itk.org/mailman/listinfo/insight-users
>
>
>
> --
> ------------------------------------------------------------------
> Simon Esneault - Therenva
> Centre d'Innovation Technologique
> Centre Cardio-Pneumologique
> CHU Pontchaillou
> Rennes, France
> Tel : +33 (0)6 64 61 30 94
> Mail : simon.esneault at therenva.com
> ------------------------------------------------------------------
>