[Vv] Fwd: Double-oblique Images in VV
Simon Rit
simon.rit at creatis.insa-lyon.fr
Fri Oct 26 11:46:04 CEST 2012
Please direct your messages to the vv list. There is a --log option in
vv to output logs to files but I have never used it, I don't know its
behavior. I will have a look at your images next week if no one
answers.
Simon
---------- Forwarded message ----------
From: Constantine Zakkaroff <mnkz at leeds.ac.uk>
Date: Fri, Oct 26, 2012 at 11:39 AM
Subject: Re: [Vv] Double-oblique Images in VV
To: Simon Rit <simon.rit at creatis.insa-lyon.fr>
Hi Simon,
I've been trying to open a couple of double-oblique files (File->Open
Image(s)) in VV and for some reason in crashes. There's a VTK error
window that pops up for a second, but it disappears too soon to read
the error message. Does VV write log files somewhere?
I've actually tried VV a few month back, but couldn't load my images
so I gave up. All of my images come from DICOM initially, but I save
them either in MHD or NRRD.
If you'd like to try and open one of the images I'm experimenting with
and the corresponding VDF, the links are below.
I've verified the CM014_angio opens just fine in Slicer 3D and in
Seg3D (http://www.sci.utah.edu/download/seg3d/), so I'm wondering what
that error message that I can't see is.
https://www.dropbox.com/s/9smifslmgic6qc6/CM014_angio.zip
https://www.dropbox.com/s/cdd28m0kv3x8h2r/CM014_vdf_perf_STRESS_15_to_4.zip
I'm able to open CM014_angio if I use File->Open nD image as (n-1)D +
t image and then I can also open the VDF. So in this case I can
pretend to examine the 3D image as a set of 2D slices. This seem to
work for the double-oblique orientation and when I use "Reset
transformation to identity".
Now, since you suggested you'd want to hear about my experiences with
VV, at this stage I'm thinking I'd like to zoom in on a part of an
image and the corresponding VDF, but I can't see a way of doing that
apart from loading a ROI from file, which I didn't try. I tried using
Tools->Crop and as a result I end up with a cropped image with the
temporal size of 1 slice, which is somewhat unexpected.
In my case I want to compare deformation fields obtained with BSpline
transforms parametrised with different grid sizes to find the optimal
grid parameters.
Overall I really appreciate you approach in emphasising the value of
deformation field visualisation and I'm going to play further with VV
to see how it can be useful in what I'm doing.
Regards,
Constantine
On 26/10/2012 09:10, Simon Rit wrote:
>
> Hi Constantine,
> It can be coarser but we don't know yet how the vector field will
> react with matrix transforms. It might work in the voxel mode but this
> is very uncertain... I would actually advise you to set the matrix
> transform to identity (right click on the transform at the bottom left
> and "Reset transformation to identity") and save your image (menu
> File, Save current image). That will then more certainly work.
> Good luck,
> Simon
>
> On Thu, Oct 25, 2012 at 2:57 PM, Constantine Zakkaroff <mnkz at leeds.ac.uk> wrote:
>>
>> Hi Simon,
>>
>> OMG. Yes, I'm sorry, I've totally missed what you said about the combo box!
>> Yes, I can see the images as expected.
>>
>> OK. I'll generate and try visualising some deformation fields. Please let me
>> ask you, with deformation fields, for example, with an N-time points 4D
>> image I can generate a N-1 frame-to-frame deformation fields if I want to
>> see the frame-to-frame deformations. Does the vector field image have to
>> match the size of the actual image, or can it have fewer frames or coarser
>> sampling?
>>
>> Many thanks,
>> Constantine
>>
>>
>>
>>
>> On 25/10/2012 13:37, Simon Rit wrote:
>>>
>>>
>>> Have you tried the combo box I mentioned in my previous email?
>>>
>>> On Oct 25, 2012 2:10 PM, "Constantine Zakkaroff" <mnkz at leeds.ac.uk
>>> <mailto:mnkz at leeds.ac.uk>> wrote:
>>>
>>> Hi Simon, hi Vivien.
>>>
>>> I must not be communicating very clearly or misunderstanding
>>> something.
>>>
>>> I've loaded a 256x256x1x76 image in VV and the screen I see is as
>>> shown in VVWindow.png attached to this message.
>>>
>>> But because the image is in double-oblique orientation (Cardiac
>>> Short-Axis plane) I'd want to see the image at least in one of the
>>> views in the as shown in CorrectOrientation.png otherwise the other
>>> views are not helpful.
>>>
>>> Is there a menu/short-cut to display the input image in
>>> double-oblique orientation by default?
>>>
>>> Many thanks,
>>> Constantine
>>>
>>>
>>>
>>> On 24/10/2012 09:39, Simon Rit wrote:
>>>
>>> Hi Constantine,
>>> You'll find here a vv snapshot from today (bin or installer)
>>> www.creatis.insa-lyon.fr/~__srit/vv.exe
>>> <http://www.creatis.insa-lyon.fr/~srit/vv.exe>
>>> www.creatis.insa-lyon.fr/~__srit/vv-1.3.0-win64.exe
>>> <http://www.creatis.insa-lyon.fr/~srit/vv-1.3.0-win64.exe>
>>>
>>> Note that the feature you want to use is largely experimental,
>>> we are
>>> working on it. You shall be able to visualize your images but
>>> there
>>> might be problems with vector fields which do not account for the
>>> transform in the image (showed at the bottom left of vv). It might
>>> still work if you slice to Voxels (top right combo box) but we
>>> haven't
>>> tested it.
>>> Could you share with us your experience on this topic? Thanks,
>>> Simon
>>>
>>>
>>> On Tue, Oct 23, 2012 at 5:20 PM, Vivien Delmon
>>> <vivien.delmon at creatis.insa-__lyon.fr
>>>
>>> <mailto:vivien.delmon at creatis.insa-lyon.fr>> wrote:
>>>
>>> Du coup il est plus intéressé par un snapshot de VV que par
>>> sa compilation
>>> ;) Si tu as le temps de faire ça.
>>>
>>> -------- Original Message --------
>>> Subject: Re: [Vv] Double-oblique Images in VV
>>> Date: Tue, 23 Oct 2012 16:08:03 +0100
>>> From:
>>> To: Vivien Delmon <vivien.delmon at creatis.insa-__lyon.fr
>>> <mailto:vivien.delmon at creatis.insa-lyon.fr>>
>>>
>>> Hi Vivien,
>>>
>>> Thanks for your reply. I'd be happy to try a snapshot if
>>> it's not too
>>> much trouble for you. I'm on MS Win Vista x64.
>>>
>>> I'd like to have a go at deformation field visualisation
>>> tool and maybe
>>> I'll discover some other relevant tools in VV.
>>>
>>> Many thanks,
>>> Constantine
>>>
>>>
>>>
>>>
>>> On 22/10/2012 15:20, Vivien Delmon wrote:
>>>
>>>
>>> On 10/22/12 14:52, Constantine Zakkaroff wrote:
>>>
>>>
>>> Hi Vivien.
>>>
>>>
>>>
>>> Hi Constantine,
>>>
>>> OK. Here's a couple links to download 2 datasets
>>> with cardiac
>>> double-oblique series.
>>>
>>> http://...
>>> http://...
>>>
>>> Please let me know if for some reason you can't
>>> download them.
>>>
>>> Would be very handy if VV was able to visualise this
>>> kind of data
>>> correctly.
>>>
>>>
>>>
>>> It can.
>>>
>>> Many thanks,
>>> Constantine
>>>
>>>
>>>
>>> I just downloaded your data and I am able to read them
>>> with the git
>>> version of vv that you can obtain with this command :
>>>
>>> git clone git://git.creatis.insa-lyon.__fr/clitk
>>> <http://git.creatis.insa-lyon.fr/clitk> vv
>>>
>>>
>>> Instructions to build vv are available here :
>>> http://www.creatis.insa-lyon.__fr/rio/vv/
>>>
>>> <http://www.creatis.insa-lyon.fr/rio/vv/>
>>>
>>> If you are not familiar with building project etc. we
>>> can build a
>>> snapshot of vv for you. We just need to know which
>>> operating system you
>>> are using.
>>>
>>> King Regards,
>>>
>>>
>>>
>>
More information about the vv
mailing list