k-Wave

1. ml414
   

   Member
   Joined: Jan '19
   Posts: 1

   Hi Dr Treeby,

   I have been using time-reversal in kwave for our real-world data and it works great. But in our experimental setup, we have a 20 mm distance of empty space (filled with coupling medium) from the transducer surface to the targeted region along the axial direction. Is there a way we can skip the recon for the empty region, but only reconstruct the targeted region so that potentially the recon speed can be accelerated (since a smaller region is reconstructed)? I've tried to cut kgrid.t_array and sensor_data, but it says t_array can only start from 0.

   Thank you!

   Posted 3 years ago #

2. bencox
   

   Developer
   Joined: Mar '10
   Posts: 292

   Hi ml414,

   It sounds as though you want to do what the seismic community call re-datuming, adjusting your data so that it is as if it were recorded at different sensor positions. If you have a planar transducer then you can use the function angularSpectrum to do that. For other array geometries it's not so straightforward, although I do recall seeing a paper, or talk perhaps, doing it with a hemi-spherical array (although frustratingly I can't find the paper now).

   Rather than use time reversal, which reconstructs the whole field for all time steps, you could reconstruct the photoacoustic image using a different type of approach. With some approaches, eg. backprojection formulas or matrix-based approaches, you could reconstruct just the region you're interested in.

   Best wishes
   Ben

   Posted 3 years ago #

RSS feed for this topic

Reply

You must log in to post.