SCIRun User Mailing List

Text archives Help


Re: [SCIRUN-USERS] basic question


Chronological Thread 
  • From: David Weinstein <dmw@sci.utah.edu>
  • To: Raya Schindmes <rayasch@medphys.ucl.ac.uk>
  • Cc: <scirun-users@sci.utah.edu>
  • Subject: Re: [SCIRUN-USERS] basic question
  • Date: Sun, 12 Mar 2006 23:04:35 -0700
Hello Raya,

This is an excellent question, and we have many of the pieces in place in SCIRun to assist you with this process.  However, there is not yet a "turn-key" solution in place.

SCIRun has several tools in place for assisting with segmentation.  Specifically, if the materials correspond to non-overlapping ranges of gray-scale values, then you can use a TransformData module to specify the material type for each range with a bit of C code.  In the more general case, where simple "thresholding" can not be used for classification, more complicated approaches are required.  In these cases, you can use the ITK modules that we have wrapped as part of the SCIRun distribution -- there is excellent documentation for the ITK filters at the ITK web-site (http://www.itk.org/).

An alternative segmentation approach that several of our users have been pursuing is to use the 3D Slicer software from the Surgical Planning Lab at the Brigham and Women's Hospital.  The steps to get Slicer up and running, and for transferring data between Slicer and SCIRun was well documented by Matthew Jolley on this page:
http://www.sci.utah.edu/ncrr/wiki/index.php/CIBC:Collab:Triedman

You asked about mesh refinement.  This is an active area of development for us on two fronts: we have been adding in support for dynamically inserting linear or surface-based models into tetrahedral finite element meshes, and we have been adding support for mesh smoothing.  The first part of this work has resulted in a new module ("InsertField"), which makes it very easy to insert and move an electrode within a volume.  This same module also supports adding contours or points onto a surface model -- that portion is still in testing, though.  Because this functionality is new, the only way to access it would be to check out a "live" version of our development tree, which is stored in "subversion".  Instructions on how to check out the code from our subversion tree can be found here:
http://code.sci.utah.edu/SCIRunAndSubversionFAQ.html
and installation directions (on building SCIRun from source) are available from this page:
http://software.sci.utah.edu/doc/

The mesh smoothing aspect of this work involves linking and taking advantage of the MESQUITE and VERDICT libraries from Sandia National Lab.  This is very much "work in progress" as well, but you are welcome to experiment with it.

As we continue to make progress in these areas, we hope to combine these various components into an easy-to-use PowerApp, complete with documentation and examples.

Cheers,
Dave


On Mar 10, 2006, at 4:01 AM, Raya Schindmes wrote:

Dear developers and users,

 

I am totally new to this software and trying to figure out what is it for exactly.

I would like to segment T1 MRI head images into 5 different tissues: white matter, grey matter, CSF, skull and scalp and then generate a tetrahedral (volume) mesh with those 5 tissues and to add electrodes position and locally refine the mesh around the electrodes.

Can anyone help me and tell me how exactly can I use SCIRun for this purpose, if at all? And what exactly should I do (preferably, step by step explanation)?

 

Thank you in advance,

 

Raya

 




Archive powered by MHonArc 2.6.16.

Top of page