Development of Connectomemapper and Connectomeviewer

[Phillip Thelen]

Hello,

As the github repositories seem rather dead I am wondering if anyone as LTS5 is still on this and/or interested in further development?

And if not, are there people interested in continuing development on a fork of connectomemapper and connectomeviewer?

greetings,

Phillip

[Alessandra Griffa]

Hi Philipp,

I'm answering you as part of the LTS5 laboratory, and particularly I'm in charge of the Connectome Mapper maintenance.
We are still interested and active in the development of the Connectome Mapper and of the Connectome Viewer.

We were silent for a while because we took time in re-structuring and re-writing particularly the Connectome Mapper, basing it on Nipype interfaces.
The rationale for this decision is that we felt the necessity of slimming the CMP code, in order to facilitate its extention, and to encourage (smoother) contributions by interested developers.

The development philosophy for the new version 2.0 of the Connectome Mapper is to keep the code of the processing as much as possible outside of the actual main Connectome Mapper code, through the use of existing Nipype interfaces and an external cmtklib library. Further development, typically additions of other tools and configuration options should go in this direction.
Note also that the GUI of the CMP 2.0 has been simplified.

The Connectome Mapper version 2.0 is currently under testing. If you wish, you can download the code at https://github.com/LTS5/cmp_nipype. We have a Wiki section with installation instructions and developers guide, while we are still updating our web-site. We would of course be very happy to receive any feed-back / comments / bugs report on the new version.

We hope that this Nipype-based CMP version will facilitate external contributions. We are currently working on this extensions:
- inclusion of reconstruction methods based on spherical deconvolution (as implemented in MRtrix software)
- inclusion of EPI distortion correction
- support for functional connectome generation (alredy present in a side branch of the 'old' CMP)

Looking forward to listening to your comments

Cheers,
Alessandra

--
You received this message because you are subscribed to the Google Groups "cmtk-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cmtk-users+...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Phillip Thelen]

Hey,

Thanks for your answer. This sounds better than I had feared.

I tried installing the Connectomemapper 2, but encountered a few problems. I opened a few issues because of that.

One problem I could not resolve (and where I'm not sure if this is a problem with my installation or connectomemapper itself) is that when I try to select one of the buttons on the left (the different steps) I get an error from traitsui, that the wx backend doesn't implement the SimpleEditor.

I am using Ubuntu 12.10 with traitsui 4.1.0. 

Through googling I found a thread that states that the SimpleEditor is only implemented for the qt backend.http://markmail.org/thread/z2rxetwqhz5lsbyj#query:+page:1+mid:dnuia225ebhvnmvt+state:results

Is it possible that connectomemapper isn't compatible with traitsui 4? And if so, it probably should be considered to switch to qt?

about contributing to connectomemapper:

I would like to help, but I'm not sure to which degree I can. Although I know python quite well, I don't really know much about MRI technology. (I'm not trying to get cmp running for me, but the researchers here at our institute.). As the Issues I opened on github all are generall problems and nothing with the processing I could try to fix them after I worked my way into the code.

greetings,

Phillip

[Phillip Thelen]

*bump*

?