k-Wave

1. huangchao
   

   Member
   Joined: Nov '10
   Posts: 57

   Hi, Dr. Cox and Dr. Treeby,

   I found that I cannot use time-reversal algorithm to exactly reconstruct images in 1D. The codes I used are as following:

   clear
   Nx = 128;
   dx = 0.05e-3;
   kgrid = makeGrid(Nx, dx);

   medium.sound_speed = 1500*ones(Nx, 1);
   medium.density = 1000*ones(Nx, 1);

   x_pos = 54;
   width = 20;
   height = 1;
   in = 0:pi/(width/2):2*pi;
   source.p0 = [zeros(x_pos,1); ((height/2)*sin(in-pi/2)+(height/2)).'; zeros(Nx - x_pos - width - 1, 1)];

   sensor.mask = [-1e-3 1e-3];

   [kgrid.t_array, dt] = makeTime(kgrid, medium.sound_speed);

   pd = kspaceFirstOrder1D(kgrid, medium, source, sensor, 'PMLSize', 20, 'PlotSim', false, 'Smooth', false);

   source.p0 = 0;
   sensor.time_reversal_boundary_data = pd;

   r = kspaceFirstOrder1D(kgrid, medium, source, sensor, 'PMLSize', 20, 'PlotSim', false, 'Smooth', false);

   Do you have idea why it's like this?

   Thank you very much!

   Best,
   Chao

   Posted 12 years ago #

2. Bradley Treeby
   

   Developer
   Joined: Mar '10
   Posts: 1,643

   Hi Chao,

   Do you mean, why do you see the extra peaks on either side of the reconstructed initial pressure? These occur because the time reversal algorithm only correctly reconstructs the pressure distribution on the interior of the measurement surface. In 1D, this means in between your two measurement points. If you truncate your field of view based on the measurement positions, the reconstructed image should agree.

   If you leave plotting on and try setting:

   'DisplayMask', cart2grid(kgrid, sensor.mask), 'PlotFreq', 1

   you can visualise what's happening.

   Note, if you use a Cartesian sensor mask, you might also notice a small numerical discrepancy when doing your comparison. This is because linear interpolation is used to calculate the pressures in the forward case, while these are enforced at the nearest grid point in the reconstruction.

   We haven't played with 1D that much, so if you notice any strange behaviour, please let us know!

   Brad.

   Posted 12 years ago #

RSS feed for this topic

Reply

You must log in to post.