SCIRun User Mailing List

Text archives Help

Re: [SCIRUN-USERS] Forward/Inverse Toolkit

Chronological Thread 
  • From: Jaume Coll-Font <>
  • To: Becky <>
  • Cc: Felipe Alonso Atienza <>, Dafang Wang <>, "Brooks.D" <>, Rob MacLeod <>, "" <>
  • Subject: Re: [SCIRUN-USERS] Forward/Inverse Toolkit
  • Date: Tue, 21 Jul 2015 14:56:55 -0400

Hi Becky,

In the beginning learning about the forward/inverse problem and how to use SCIRun can be a bit tricky. Don't worry about asking questions.

Are you trying to use one of the networks in the toolkit? If so, you may be having problems with the path to the data associated with SCIRun. To solve this, you need to go to "Configure" in the lower right corner of SCIRun and change the path of "SCIRun Data" in the "Options" tab.

This should allow you to run the default configuration of any example network.

The question of what is being loaded in each readMatrix module, depends on the network you are testing.
In general, the inverse networks will have the following inputs:
- forward matrix
- heart potentials
- body surface potentials (note that to provide well easy to understand examples we "cheat" and synthetically propagate the potentials on the heart to the body surface. In this case the body surface potentials are not being loaded but are the result of a matrix multiplication + noise)

Then, the results can be plotted on the heart mesh that needs to be inputed through a ReadField module.
THere could also be a torso geometry to plot BSPM.
The names in of the input matrices should be clear about what they are. For example a forward matrix should have "fwd" or "transfer matrix" in it or a heart surface potentials should have "pots" or "egm".

Does this help?


Jaume Coll-Font

On Tue, Jul 21, 2015 at 2:37 PM, Becky <> wrote:
Thank you all so much for the help so far. I have read through the documentation and it is helpful in showing which modules to use in SciRun, however I am still struggling with which matrices to select at each ReadMatrix module. I am very new to the whole concept of forward/inverse solution, and apologize for my lack of basic knowledge here... Basically I have been given some sample data, including meshes, time signals, and some transforms (which I know were output FROM SciRun to begin with), and I am trying to use these to gain an understanding of how to use SciRun before I try to build my own meshes and use my own collected body surface ECG signals.

If anyone is able to show me which kind of files they input into the modules along with the schematic of modules in SciRun, I would really appreciate it.

Thanks so much.

On 21 July 2015 at 02:20, Felipe Alonso Atienza <> wrote:
Hi beck, 

At the end of this document you will find some examples on how we connected the different modules providing a body surface, a heart surface and a transfer matrix: 

Hope it helps.


Felipe Alonso Atienza

Universidad Rey Juan Carlos
Dept. Teoría de la Señal y Comunicaciones
D208 Departamental III
Camino del Molino s/n, 28943 Fuenlabrada, Madrid

Tlf:   91 488 8702. Fax: 91 488 7500
Twitter: @FelipeURJC

El 21/7/2015, a las 1:32, Rob MacLeod <> escribió:


You should be taking out IBBM Summer course.  We will discuss this on Thursday. (-:

Next year, perhaps?  

I am sure some others on the list can provide some specifics in the meantime.


On Jul 20, 2015, at 16:46 , <> wrote:

Hi there,

I'm trying to get familiar with SciRun, and struggling to properly use the
Forward/Inverse Toolkit. I'm trying to use it ECG taken across the body
surface and using the inverse solution to visualize the signal on the heart

Do you have any examples of a functioning inverse solution? I hope this would
help me figure out what kind of files are required/ what file format the data
should be in.


Archive powered by MHonArc 2.6.18.

Top of page