[Nipy-devel] Latest and greatest in visualization of MRI data?

Yann Cointepas yann@sapetnioc....
Tue Jan 25 02:43:08 CST 2011

There are two implementations of the Anatomist Python API.

1) The "socket" one where Anatomist runs in its onw process and listen to
commands on a socket. The API convert Python code to Anatomist commands and
send them to the socket. The main advantage of this implementation is the
ability to command a remote Anatomist eventually running on a different OS.
For instance, with a few line of Python code Denis Rivière (the Grand Guru
of Anatomist) connected the orientation of a displayed brain to the
orientation of an Android smartphone.

2) The "direct" one where C++ code is directly called via a Python wrapping
generated using Sip (the PyQt C++/Python bindings generator). To use this
API you must start your own Qt event loop (it works perfectly well with
ipython -q4thread). This API has more features since you can have a direct
access to Anatomist objects from Python. For instance you can use Numpy to
modify an image displayed by Anatomist or you can embbed Anatomist widgets
in your own GUI.


Yann Cointepas            Tel: +33 1 69 08 78 31
CEA - Neurospin           Fax: +33 1 69 08 79 80
Bâtiment 145, Point Courrier 156
91191 Gif-sur-Yvette cedex, France

On Mon, Jan 24, 2011 at 9:55 PM, Gael Varoquaux <
gael.varoquaux@normalesup.org> wrote:

> On Mon, Jan 24, 2011 at 09:53:31PM +0100, Yann Cointepas wrote:
> >    Anatomist is entirely controlable from Python. For instance the
> following
> >    sample script can be used to load, fusion and display two images.
> >  # Initialize Anatomist API
> >  import anatomist.api as anatomist_api
> Good point!
> How does Anatomist deal with mainloop issues: does it start its own (Qt)
> mainloop, or can it use an existing one, as when IPython is started with
> the '-qthread' option?
> Gaël
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.scipy.org/pipermail/nipy-devel/attachments/20110125/769b848b/attachment.html 

More information about the Nipy-devel mailing list