I am in the process of registering Mindboggle (http://www.binarybottle.com/mindboggle.h...) with NITRC, and am about to completely rewrite Mindboggle to have a more efficient, flexible and modular processing pipeline. I am strongly considering porting the codebase over to Python for the following reasons:

1. free, open source (unlike Matlab)
2. easy to program, understand, and maintain
3. easy for Matlabbers (e.g., python's Matplotlib)
4. recent development of numpy (scipy.org)
5. widely used

As evidence of #5, beyond the recent effort of NiPy (http://neuroimaging.scipy.org/) to create a "full-featured python program and development environment for analyzing functional imaging data," and the extensive list of scientific programs on scipy.org's website, I'm finding that many software suites are now using Python as the scripting language that "glues" pieces of code together (e.g., ESRI's GIS software).

I would be very interested to hear what people have to say about porting code over from Matlab to Python for use by the neuroimaging community.

Thank you!

Cheers,
@rno