Hi Arpit, you are welcome.
Just two hints. 1) It is a multimodal problem. So look for multimodal cost functions like Normalized Mutual Information or Mattes Mutual Information (in ITK, mostly the search space is less noise then with the simple itk Mutual Information). 2) From your mail I assume that you have a intra-personal problem statement (aka registering data of one person and not of many persons). If it is just about registering the Heads (skull) then a rigid transformation model should be sufficient. If you have changes in the brain topology (e.g. brain shift due to surgery or tumor growth, changes in longitudinal follow up) which you want to compensate it becomes much harder (and a deformable registration problem). Best regards, Ralf Von: Arpit Paliwal [mailto:[email protected]] Gesendet: Freitag, 27. Februar 2015 15:01 An: Floca, Ralf Omar Cc: [email protected] Betreff: Re: CT MR 3D Fusion (Arpit Paliwal) Hi Ralf, Thanks for your detailed reply. I am new to image based applications. I would like to do CT & MR registration of human brain with C++ classes. Thanks for pointing me to ITK. It seems ITK framework can help me with the registration. In the meantime I'll also explore the working of MITK workbench for the registration. Best Regards, Arpit On Wed, Feb 25, 2015 at 9:49 PM, Floca, Ralf Omar <[email protected]> wrote: Dear Arpit, regarding your questions > Is there any class available in MITK to achieve the 'automatic' 3D > registration of CT & MR ? Just to be sure that we are not talking on crossed purposes: Are you searching for a solution on (1) the C++-class levels (so e.g. for own programs) or on (2) the MITK workbench level? For (1) I would advise you to look at the registration framework that is part of ITK [1] and offers several options for image registrations. Have also a look at the ITK SoftwareGuide book 2 chapter 3 "Registration" [2] or at the ITK mailinglist. The ITK registration pipeline - as any itk filter pipeline - can directly be integrated in your MITK code. For (2) There is a registration plug-in in the MITK release which offers a wrapping for ITK v3 registration algorithms. So there is the possibility to determine rigid/affine registrations. Remark 1: to exploit all possibilities the advanced tab must be activated. See a recent mailinglist conversation on that topic [3]. Remark 2: The current implementation in the MITK workbench only produces the mapped image. So on this level you have no access the transformation determined by the registration algorithm > registration of CT & MR ? Despite the technical level above, the answer is: It depends. It depends heavily on your problem statement (e.g. which body region, quality/peculiarities of your input data, what is the goal of your registration (it makes a huge differences e.g. if you look at the pelvis for example and you say, I want the bony structures registered or the bowel), and so on). So this question cannot be answered reasonably without knowing more details. Hope this answer helps at least a little bit. Kind regards, Ralf Floca [1] http://www.itk.org/ [2] http://www.itk.org/ItkSoftwareGuide.pdf [3] http://sourceforge.net/p/mitk/mailman/message/33466315/ > -----Ursprüngliche Nachricht----- > Message: 3 > Date: Sun, 15 Feb 2015 22:07:06 +0530 > From: Arpit Paliwal < > > Subject: [mitk-users] CT MR 3D Fusion > To: "[email protected]" > <[email protected]> > Message-ID: > <CALTt7XiivY- > [email protected]> > Content-Type: text/plain; charset="utf-8" > > Dear All, > > I am working on 3D CT / MR image registration (fusion). > Is there any class available in MITK to achieve the 'automatic' 3D > registration of CT & MR ? > For e.g. I provide CT & MR images as input & get the required > transformation as output. > > Best Regards, > Arpit > -------------- next part -------------- > An HTML attachment was scrubbed... > > ------------------------------ > > ------------------------------------------------------------------------ > ------ > Dive into the World of Parallel Programming. The Go Parallel Website, > sponsored by Intel and developed in partnership with Slashdot Media, is > your hub for all things parallel software development, from weekly > thought leadership blogs to news, videos, case studies, tutorials and > more. Take a look and join the conversation now. > http://goparallel.sourceforge.net/ > > ------------------------------ > > _______________________________________________ > mitk-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/mitk-users > > > End of mitk-users Digest, Vol 105, Issue 9 > ****************************************** ------------------------------------------------------------------------------ Dive into the World of Parallel Programming The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net/ _______________________________________________ mitk-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/mitk-users
