I need to start a project which deals with the DICOM data manipulation, visualization. As a part of some basic research, I have found there are a few toolkits such as ITK, VTK which performs data manipulation on the medical image data. My question is, using ITK+VTK+QT for DICOM image manipulation(segmentation & registration) a better choice or using ITK with OpenCV would be a better option? Or any alternate toolkits exists that would likely achieve my requirement?
Any suggestions, sources or links related to this topic would be much helpful.
There are many toolkits and frameworks that work with Dicom. It depends on what you want to do.
In many cases, the easiest thing to do is to build a plugin for an existing application/toolkit such as Horos, 3D Slicer, ImageJ, MITK, MeVisLab, ITK Snap, etc. I'm sure there are dozens more.
If you want to build your own medical imaging application, most of the above are open-source; adapting one of these would save you a lot of grief (and probably years) compared to trying to write your own application from scratch.
If your main interest is in developing algorithms, then Python is a good prototyping language - consider packages such as numpy, scipy, pydicom, ITK, SimpleITK. Java has dcm4chee. C++ has QT, ITK, VTK.
If you want to do something JavaScript-based that will work though web browsers on tablets etc, look at in-progress projects such as OHIF Viewer or Cornerstone.
One other thing: a) dealing with Dicom data, b) manipulation and c) visualisation, are three different things. It's easy to convert your Dicom data to, say, nifti format, which opens up a lot of academic analysis tools. Similarly, there are many 2D and 3D visualisation libraries that are not specific to Dicom - it's just about converting data into the right form.