It's not my specific area of expertise but I can direct you to some who work on it:
Methods: yes. There is tons of literature out there that gives methods to do this. Just do a google scholar search on "tumor segmentation". Or check this researcher's publications: https://sites.google.com/site/marcelprastawa/
Implementation: If you want a specific implementation in, say, Matlab you can google some code. Eg: http://www.mathworks.com/matlabcentral/answers/78776-how-segmenting-brain-tumor-using-matlab-code.
More links: Lots of researcher's are specifically working on segmentation of MRI images etc, using computational methods. Most of them use Matlab. You can directly contact students/researchers working for
http://www.sci.utah.edu/~gerig/publications.html
to ask if they have specific open source implementation in Matlab and I'm sure they'll be happy to help.
I am closely working in this field,although it is not the specific topic of my research. Anyway, I first tell you that it depends. It depends on(some examples):
- Which is the location of the tumor. According to this(prostate, lung, brain..), image modalities used may be different. (PET,MRT1,T2...)
- Which kind of tumor. For instance, in brain tumors, does it infiltrate? or does it push to surrounding tissue?
- Which kind of user interaction? The approach could be different if you want a semi-automatic or a fully automatic method.
As examples:
In lung cancer tumor(non-small cell lung cancer), I have seen methods like: threshold based(rule of 40%), random walkers,fuzzy cluster or graph cuts.
In brain cancers, also interactive approaches like graph cuts can be applied. Or fully automatic methods like machine learning approaches.
Having said that I think that you will be able to find the matlab implementation of most of these or other techniques.