Import metabolites map from NMRspectroscopy

Questions about BrainVisa usage and installation

Moderators: denghien, riviere

Post Reply
bagory
Posts: 7
Joined: Tue May 23, 2006 1:11 pm
Location: CERMEP, Lyon, France

Import metabolites map from NMRspectroscopy

Post by bagory »

Hello,

I’m working on NMR spectroscopy (on a Siemens Sonata 1.5), and I would like to use Brainvisa to import 2D metabolite maps and fusion them with brain T1 WM/GW/CSF segmented data.

I obtain this map files from a spectral quantification software (jMRUI). It’s mainly consist of a text file of one “value” per metabolite and per voxel (of ~1 cm^3).

My goals are to develop a process which import my metabolite maps, store them in the BrainVisa database on a appropriate format (to define ?) for further treatments, run a segmentation process on T1 anatomy data, and build fusions map (with interpolation and smoothing) in Anatomist between T1 segmented images and metabolites maps.

My questions are :
1) Do you know if a spectroscopy team has already develop such a process in BrainVisa ?
2) Is such a project seems realist and feasible to you ?

Thanks in advances for your advices,

Matthieu
Manik
Posts: 99
Joined: Fri Mar 05, 2004 1:00 pm
Location: INSERM U836, Grenoble

Post by Manik »

Hi,

I do not know if someone is working on this kind of process, but I can tell you that your project is realistic and feasible.
I think you could convert your data to a format like the graph format of anatomist (.arg file with associated data). I am not an expert about this, but I think you can put a 4D image (GIS format, very easy to write, composed of a .ima and a .dim file) as the data, which would be a 3D image for each metabolite containing the measured value, the 4th dimension being the different metabolites. The .arg file allows you to give a label to each of these 3D volumes (for example the name of the metabolite).
Others will be able to give more detailled information about the way to do this, for example using the C++ libraries of Brainvisa/Anatomist.
If the voxel size is available in your files, it will be easy to put this information in the .dim file generated, so Anatomist will know the voxel size and will be able to fuse your image with a T1 MRI.

Smoothing/interpolating an image can be done with Anatomist but I am not sure how Anatomist would handle this in a Fusion. To get rid of this problem, you can call AimsResample from your process to put your NMR image at the same spatial resolution as your T1 MRI (interpolation), then tell Anatomist to fuse the images.

Good luck !
Manik Bhattacharjee
INSERM U836
bagory
Posts: 7
Joined: Tue May 23, 2006 1:11 pm
Location: CERMEP, Lyon, France

Post by bagory »

Hi Manik,

Thanks a lot for your advices. I will tell you if it works.

Matthieu
Post Reply