[Insight-users] How to set the fixed image region in the registration framework v4 ?

[Champ Frédéric]

Hi Tim,

- In your case the thread must be set to 1 as the* non-v4 version seems 
to be not thread safe*.
- My problem is different: I want a thread safe slice-by-slice 
registration method by using the registration frame work v4 but I *I 
can't specify the fixed image region as a slice without an execution 
error* (e.g. my previous message).

Fred

Le 21/01/2013 17:58, Tim Allman a écrit :
> This is interesting. I get the same problem with the non-v4 version of 
> MeanSquaresImageToImageMetric but it goes away if I set the number of 
> threads to 1. It might be worth a try to see if this cures your 
> problem too.
>
> Tim
>
> On 13-01-21 03:58 AM, Champ Frédéric wrote:
>> Hi,
>>
>> I'm working on a slice-by-slice registration method for diffusion 
>> MRI.  The method works with the registration framework but the  
>> itk::MattesMutualInformationImageToImageMetric (which gives best 
>> results) is not thread safe.
>>
>> I tried to optimize this method be using the framework v4 but I get 
>> some problems.
>> To compute a slice-by-slice registration, In the old framework I set 
>> the current slice as the fixed image region at each iteration. In the 
>> new framework, this region is used to create the mask of fixed image 
>> set in the metric (It's the only way I found to specify the image 
>> region...) but it doesn't work and returns the following message:
>>
>> Description: itk::ERROR: 
>> MattesMutualInformationImageToImageMetricv4(0x7f7eb004b360): *Too 
>> many samples map outside moving image buffer. There are only 4 valid 
>> points out of 471 total points*. The images do not sufficiently 
>> overlap. They need to be initialized to have more overlap before this 
>> metric will work. For instance, you can align the image centers by 
>> translation.
>>
>> However if I don't set any mask the registration works but it 
>> registers the whole volume...
>> I tried to use the same Image as the fixed and moving one (expecting 
>> an overlap) or to use fixed sample point sets but in both cases I had 
>> the same error message.
>> Did I missed some step ? Is there any way to specify the fixed region ?
>>
>> Thanks,
>>
>> Fred
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> _____________________________________
>> Powered bywww.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
>>      
>
> -- 
> Tim Allman, Ph.D.
> 35 Margaret Street,
> Guelph Ontario N1E 5R6
> Canada
>    
>
>
> _____________________________________
> 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
>    

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20130121/85b48c8c/attachment.htm>