k-Wave

1. huangchao
   

   Member
   Joined: Nov '10
   Posts: 57

   Hi,

   It seems that the time-reversal reconstruction is not available in the current release of the C++ code. Is this right? I just want to confirm.

   Thanks,
   Chao

   Posted 12 years ago #

2. Bradley Treeby
   

   Developer
   Joined: Mar '10
   Posts: 1,643

   Hi Chao,

   The original method of using time-reversal (i.e., setting sensor.time_reversal_boundary_data) is not included in the C++ code. However, you can use time-reversal in the more general way of reversing the recorded boundary data and applying it as a Dirichlet boundary condition. If you look inside the function kspaceFirstOrder3DC, you can see an example of how this is done. The net result is the same via both methods.

   Brad.

   Posted 12 years ago #

3. bernhardkaplan
   

   Member
   Joined: Sep '15
   Posts: 1

   Hi,
   I would like to use k-wave's C++ code on a compute cluster without matlab via the command line and to do time-reversal reconstruction instead of FFT, but I am not able to create the HDF5 files using kspaceFirstOrder3D(kgrid, medium, source, sensor, 'SaveToDisk', filename);

   When looking at kspaceFirstOrder3DC.m time_rev is set to true if the sensor has a time_reversal_boundary_data field -- however, when calling the kspaceFirstOrder3D function, an error is raised with the message "The optional input 'SaveToDisk' is currently only compatible with forward simulations using a non-zero binary sensor mask", but the sensor mask is non-zero.

   Could you explain in more detail how to use the C++ for time-reversal reconstruction instead of FFT? An example script would be fantastic (I was using the example_pr_3D_TR_planar_sensor.m as starting point so far).

   Many thanks in advance and best regards,
   Bernhard Kaplan

   Posted 9 years ago #

4. Bradley Treeby
   

   Developer
   Joined: Mar '10
   Posts: 1,643

   Hi Bernhard,

   As mentioned above, you can use the C++ code directly for time reversal, but you need to manually reverse your recorded boundary data, and then use this as a Dirichlet source in your simulation. In steps:

   (1) Instead of assigning your boundary data to sensor.time_reversal_boundary_data, reverse it in time and assign to source.p, i.e.,

   source.p = flipdim(boundary_data, 2);

   (2) Set the source to be a Dirichlet boundary condition

   source.p_mode = 'dirichlet';

   (3) Instead of assigning the position of your detectors to sensor.mask, assign it to source.p_mask

   (4) Set sensor.record = {'p_final'} to record the final (i.e., reconstructed) pressure field. The sensor mask isn't used in this case, so anything will do. Your reconstructed image will be contained in the field sensor_data.p_final

   If you are stuck on how to get started with time-reversal using the normal syntax, there are several examples included in the toolbox.

   Hope that helps,

   Brad.

   Posted 9 years ago #

RSS feed for this topic

Reply

You must log in to post.