[Insight-users] about itkNormalizedMutualInformationHistogramImageToImageMetric and OPEN ACCESS

[Luis Ibanez]

Hi Yanqiu,

You are right, the HistogramImageToImageMetrics
use all pixels in the image in order to compute
the histogram.

There is currently no option for you to select
only a random sampling of pixels...

BUT....

Here is where you have the pleasure of using
Open Source Software !!!  :-)

You can change your own copy by yourself !



Just go to the file:

   Insight/Code/Algorithms/
      itkHistogramImageToImageMetric.txx

and replace the ImageRegionIterator with the
RandomImageIterator.


For details on the use of Image Iterators,
please read the ITK Software Guide:

   http://www.itk.org/ItkSoftwareGuide.pdf


where you will find a Chapter on "Image Iterators"



Welcome to Open Source !


BTW, if you find that replacement useful, (or even
if you don't) you are strongly encouraged to report
your results to the Insight Journal.

http://www.insightsoftwareconsortium.org/InsightJournal/

In the Insight Journal, your results should be posted
along with the source code that you used for testing
and the input images and output image that you obtained.

The Insight Journal will attempt to replicate your
results using an automatic Dashboard. You are welcome
to post papers (plus source code and data) that describe
the replication of previous work, regardless of whether
you obtain positive or negative results.

Real scientific Journals must publish papers on:

* Replication of work published by others.
* Reports of Negative results.
* Corrections to previous work.

and of course, in order to exercise the scientific
method, the work described in papers *MUST* be
reproducible by others, which in Image Processing
means that you must post your source code, your
images and your parameters.

Journal that do not support nor encourage these fundamental
characteristics of the scientific methods are simply "Vanity"
Journals, where researchers submit papers just to put them in
their CV and in their annual reports.

The corrupted and narrow mentality of "Publish or Perish"
have unfortunately ruined the scientific education of two
generations of researchers, making them believe that the
purpose of research is "to Publish".

The Open Access movement is here to correct this shameful
situation.

Young (in age or spirit) researchers can still be saved
from pursuing pointless careers !!


More information about the Open Access revolution can be
found at


      http://en.wikipedia.org/wiki/Open_access





  Regards,



     Luis



----------------------
Yanqiu Feng wrote:
> Dear to All,
>  
> For 3D data, I found that registration with the 
> itkNormalizedMutualInformationHistogramImageToImageMetric or 
> itkMutualInformationHistogramImageToImageMetric takes a very long times 
> even we give the small size of the histogram. This is because that it 
> transforms and interpolates all points in the Fixed image region in 
> order to compute the metric at each iterattion.
>  
> Is there a method to set the number of spatial samples in a similar way 
> to MattesMutualInformationMetric?
>  
> Best Regards.
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Insight-users mailing list
> Insight-users at itk.org
> http://www.itk.org/mailman/listinfo/insight-users