k-Wave

1. rmurthy
   

   Member
   Joined: Nov '11
   Posts: 4

   Hello,

   I am able to set medium.density and medium.sound_speed as matrices, but I encounter errors when trying to set medium.alpha_coeff and medium.alpha_power to matrices... is there a way to circumvent this? I'm trying to simulate ultrasound propagation through two adjacent tissues with very different attenuation coefficient values, so I'd like to be able to define all of those parameters to reflect that.

   Thanks!

   Posted 13 years ago #

2. Bradley Treeby
   

   Developer
   Joined: Mar '10
   Posts: 1,643

   Thanks for your feedback. There has also been some other discussion of this issue on the forum. In the B.03 release the absorption values must be homogeneous, however, this has been modified in B.04 to allow a heterogeneous medium.alpha_coeff term.

   Posted 13 years ago #

3. mgt_p
   

   Member
   Joined: Jul '13
   Posts: 10

   Hello,

   Just to check, having BonA and alpha* as matrices as not been implemented yet, right?

   Thanks

   Margot

   Posted 10 years ago #

4. Bradley Treeby
   

   Developer
   Joined: Mar '10
   Posts: 1,643

   Hi Margot,

   Both medium.alpha_coeff and medium.BonA can be defined as heterogeneous matrices. One thing to keep in mind is that if you set B/A to zero, this will turn off the material nonlinearity, but there will still be a convective nonlinearity. In other words, setting the nonlinear parameter B/A to zero is not the same as setting the coefficient of nonlinearity beta = 1 + B/2A to zero. If you want to do the latter and you are using the MATLAB code, you can modify the simulation functions to do this. For example, using V1.0 in 2D, open kspaceFirstOrder2D and replace lines 808 and 809 with:

   % force the convective nonlinearity term to be zero in regions of the domain where BonA is zero
   rhox = bsxfun(@times, pml_x, ( bsxfun(@times, pml_x, rhox) - dt.*rho0 .* duxdx ) ./ (1 + (medium.BonA ~= 0) .* 2*dt.*duxdx));
   rhoy = bsxfun(@times, pml_y, ( bsxfun(@times, pml_y, rhoy) - dt.*rho0 .* duydy ) ./ (1 + (medium.BonA ~= 0) .* 2*dt.*duydy));

   Hope that helps,

   Brad.

   Posted 10 years ago #

5. MO_2015
   

   Member
   Joined: Nov '15
   Posts: 1

   Hello,

   I got a question regarding medium.alpha_power which is not really answered above. Is it possible to define this parameter as a heterogeneous matrice just like medium.alpha_coeff?

   Thanks in advance!

   Kind regards
   MO

   Posted 8 years ago #

6. benji
   

   Member
   Joined: Dec '15
   Posts: 14

   hi,
   first - I have exactly the same question of MO above - can medium.alpha_power be defined as a heterogeneous matrix?

   secondly - I have tried defining different homogeneous medium.alpha_power and medium.alpha_coeff and run a simple simulation - I keep getting exactly the same pressure profile in the sensor, independent of the loss defined (in other words - seams like the loss parameters do nothing...)

   How Is that possible?

   Thank u,
   benji

   Posted 8 years ago #

7. Anthony
   

   Member
   Joined: Apr '13
   Posts: 96

   Hi benji,

   Unfortunately, alpha_power cannot be heterogeneous because it is involved in the definition of the fractional differential operators used for the absorption... (see the paper "Modeling power law absorption and dispersion for acoustic propagation using the fractional Laplacian." for the formula).

   I don't know if there would be ways to circumvent this limitation but, with the code as it is, I don't think it is possible.

   About your second question, could you post a code?

   Best regards,
   Anthony

   Posted 8 years ago #

8. Anthony
   

   Member
   Joined: Apr '13
   Posts: 96

   Hi benji,

   I found the discussion I had with Brad on this particular topic:
   http://www.k-wave.org/forum/topic/heterogeneous-alpha_power

   Best regards,
   Anthony

   Posted 8 years ago #

RSS feed for this topic

Reply

You must log in to post.