k-Wave

1. HannaB
   

   Member
   Joined: Sep '17
   Posts: 5

   Hello,

   I have a 3d simulation working fine with kspacefirstorder3D, but taking few hours to complete. I am trying to improve the code by running it on my GPU.
   I only change on my code the function - replaced by kspacefirstorder3DG.
   I don't understand how to control what happens inside this function which seems to do something different than before.

   - 'kspacefirstorder3D' code gives me sensor data of 64 (elements) by 4468 (time)

   - 'kspacefirstorder3DG' code gives me sensor data of 3200 by 4468. By plotting the RF I understood where the 3200 comes from:
   3200 = 64 (elements) * 5 (size of element in kgrid) * 10 (repetitions of the simulation)

   Here are my points :
   1) Why is the simulation repeated 10 times? How can I ask for only one?
   2) Why do I have 5 signals per element and not only one? Can I change this in the inputs? (I guess I can fix this by summing the 5 signals for each element, but how come it is different from kspacefirstorder3d?..)

   Thank you very much for your answers and advice,
   All the best,

   Hanna

   PS: The transducer is defined with kWaveTransducer.

   Posted 7 years ago #

2. HannaB
   

   Member
   Joined: Sep '17
   Posts: 5

   Update: I think I understood the point but still have questions.

   The width of one element is 5 in grid units, and its length is 10 in grid units. The output signal given by kspacefirstorder3dg seems to give the RF coming from each grid point composing each element of the transducer. Is that right?

   Is there a way to ask for only 64 signals in the output? Or do I have to code the summation for each element of my transducer myself?

   Therefore, is it a summation of the RF from each grid points of the element that is done in kspacefirstorder3d? Or is it a mean?

   Thanks for your answers,

   Hanna

   Posted 7 years ago #

3. Bradley Treeby
   

   Developer
   Joined: Mar '10
   Posts: 1,643

   Hi Hanna,

   When you use an object of the kWaveTransducer class as the sensor, the MATLAB implementation of kspaceFirstOrder3D will automatically average the signals across the grid points belonging to each transducer element using the appropriate delays. The output sensor_data will then contain as many time series as there are transducer elements.

   This behaviour is not implemented in the C++ code, so instead the output sensor_data will contain as many time series as the number of grid points that form the transducer. However, you can combine the data after the simulation using the combine_sensor_data of the kWaveTransducer class to give you the same output as when using the MATLAB code.

   Brad.

   Posted 7 years ago #

4. HannaB
   

   Member
   Joined: Sep '17
   Posts: 5

   Hi Brad
   Couldn't have been more helpful, thanks a lot !!

   Posted 7 years ago #

RSS feed for this topic

Reply

You must log in to post.