[Insight-users] statistics on two groups of image volumes

[Daniel Mace]

Dennis,

I'm also not entirely sure you would like to do with the two groups of 
images.  So you are looking for a filter that would take as input these 
two groups of images and return... differences within the pixels 
regions?, differing variability between the two groups? different 
changes between regions?  There are a lot of questions that one could ask.

I'm guessing by your email extension that you would probably be 
interesting in applying something along the lines of Statistical 
Parameteric Mapping (or some kind of modeling on the pixel level) 
between the two groups of images-- which, to my knowledge, doesn't exist 
in ITK (someone chime in if I'm wrong on this one).

I'm not sure there is an exact filter that would specifically take two 
groups of images and return a vector of homologous pixels.  The closest 
and easiest solution would be to convert your separate images into a 
VectorImage and use an iterator to extract out the vector of homologous 
pixels (see ImageToVectorImageFilter ( 
http://www.melaneum.com/OTB/doxygen/classitk_1_1ImageToVectorImageFilter.html 
), if you iterate over the output of that filter, your pixel types will 
be a vector of your homologous pixels).  You'd either have to use two of 
them to grab two groups of vector images, or just use one and have a 
small vector that defines the groups.

Cheers,
Dan

Dennis Jen wrote:
> Hi,
>  Does anyone have any advice on this?
>  Thanks,
>    Dennis
>
> Dennis Jen wrote:
>> Hi Luis,
>>  I'd like an arbitrary number of image volumes ideally, and I imagine 
>> between 10-100 volumes as input.  The images will be registered and 
>> the same size.  My ultimate goal is to write a statistics filter that 
>> would be applied to the homologous pixel across two groups of image 
>> volumes, so the two collections of image volumes would be input.  I 
>> was hoping that visiting homologous pixels across a group of image 
>> volumes and applying some sort of strategy might already be 
>> encapsulated within a filter.
>>  Please let me know if I can provide more details.
>>  Thanks,
>>    Dennis
>>
>> Luis Ibanez wrote:
>>> Hi Dennis,
>>>
>>> Could you please give us more details on what you are trying
>>> to do.
>>>
>>>
>>> I would assume that you have a collection of images and that...
>>>
>>> you want to compute the average intensity of all the images ?
>>>
>>> or the standard deviation of the intensities of the homologous
>>> pixel across the images ?
>>>
>>> Are the images of the same size ?
>>>
>>> Are the images registered ?
>>>
>>> Are you talking about 2 images ? 10, 200, 2000 ?
>>>
>>>
>>> Depending on the scenario that you have in mind, different
>>> options available in ITK could apply.
>>>
>>>
>>> Please give us a more detailed description of what you
>>> are attempting to do.
>>>
>>>
>>>     Thanks
>>>
>>>
>>>       Luis
>>>
>>>
>>> ------------------
>>> Dennis Jen wrote:
>>>> Hi,
>>>>  Is there a class or example for operating over the pixels of an 
>>>> arbitrary (perhaps a vector) of image volumes?  I'd like to 
>>>> ultimately apply a statistics test on two arbitrary-sized 
>>>> collections of image volumes.
>>>>  Thanks,
>>>>    Dennis
>>>> _______________________________________________
>>>> Insight-users mailing list
>>>> Insight-users at itk.org
>>>> http://www.itk.org/mailman/listinfo/insight-users
>>>>
>>>
>>>
>> _______________________________________________
>> Insight-users mailing list
>> Insight-users at itk.org
>> http://www.itk.org/mailman/listinfo/insight-users
>>
>>
> _______________________________________________
> Insight-users mailing list
> Insight-users at itk.org
> http://www.itk.org/mailman/listinfo/insight-users