k-Wave

1. zhuoxiaoyu
   

   Member
   Joined: Nov '17
   Posts: 5

   I'm new in K-Wave, now I need to do a simulation about proton beam and proton-induced acoustic emission.

   In the simulation, The position of the Bragg peak will change with the energy of the proton beam. Now I already have a function about the dose distribution of the proton beam and Bragg, different dose values will produce different levels of acoustic emission.

   but I don't know how to simulate the acoustic emission images of the proton beam (including the Bragg peak) by using the function "kspaceFirstOrder3D".I've finished defining grids, medium, and sensors. however, I don't know how to use "source.p0"(or source.p_mask and source.p?) to express the proton beam and its acoustic emission.

   I want to get some help, thank you!

   Posted 6 years ago #

2. Bradley Treeby
   

   Developer
   Joined: Mar '10
   Posts: 1,643

   Hi zhuoxiaoyu,

   I would suggest using source.p_mask to define the spatial distribution of the source, and then use something like a Gaussian shaped source.p with the appropriate width to model the energy deposition. You will need to define source.p as a matrix, with the amplitude scaled to the deposited dose at each spatial position. See this paper for more details.

   Brad.

   Posted 6 years ago #

3. HuiLIN
   

   Member
   Joined: Nov '17
   Posts: 7

   Dear Brad.,

   Could you speak more about how to describe a matrix source.p with different amplitude more detail? Is there appropriate example in k-Wave examples for ref?
   My source is an inhomogeneous spatial distribution with a periodic input_signal. When I use
   source.p(i,j,k) = source_strength(i,j,k).* input_signal;
   to define the source, the code always breaks for different dimension or “Index beyond the matrix dimension”.
   Thank you a lot in advance.

   Hui

   Posted 6 years ago #

4. Bradley Treeby
   

   Developer
   Joined: Mar '10
   Posts: 1,643

   Hi Hui,

   The source must always be a two-dimensional matrix with dimensions (num_source_points, num_time_points). The source points are indexed according to MATLAB's standard column-wise linear matrix index ordering.

   Brad.

   Posted 6 years ago #

5. zhuoxiaoyu
   

   Member
   Joined: Nov '17
   Posts: 5

   Hi Brad.,

   When I use

   for i=1:Nx
   for j=1:Ny
   source.p(i,j)=mag_distr(i,j);
   end
   end

   to define the source, the simulation couldn't start for "Input signal must be a vector".
   ( mag_distr(i,j) represents pressure distribution about proton beam dose deposition

   Could you tell me how to define source.p as a matrix ?

   Posted 6 years ago #

6. zhuoxiaoyu
   

   Member
   Joined: Nov '17
   Posts: 5

   I think I've probably figured out what (num_source_points, num_time_points) means.

   I modified my code, but I still had some problems.

   When I use

   x = 0:1e-7:1e-5;
   input_signal = gaussmf(x,[5e-7 8e-6]);
   source.p(1:sum(source.p_mask ~= 0), :) = source.p_mask(source.p_mask ~= 0) * input_signal;

   to define the source, the simulation breaks for "Subscripted assignment dimension mismatch".

   Could you tell me how to solve this and how to assign "mag_distr(i,j)"(Two dimensional matrix of the same size as the grids) to "source.p"?

   Thanks a lot!

   Posted 6 years ago #

7. bencox
   

   Developer
   Joined: Mar '10
   Posts: 292

   Hi Hui,

   Each row of the matrix source.p should be a time series.
   Each row (each time series) is the source applied at one source point in source.p_mask.

   The source points are ordered columnwise, ie. the first time series (the first row of source.p) is the source applied at the source point with the smallest index according to Matlab's usual matrix numbering convention:
   https://uk.mathworks.com/help/matlab/math/matrix-indexing.html, and so on.

   Because each row is associated with one grid point, it can be weighted to give the source at that point a different amplitude. The weighting will be determined beforehand, in your case using Geant4 or similar.

   Hope that's clear!
   Ben

   Posted 6 years ago #

8. HustYjq
   

   Member
   Joined: Jul '20
   Posts: 20

   Hi,zhuoxiaoyu
   Have you worked out this question?I have the same problem with you.

   Posted 4 years ago #

RSS feed for this topic

Reply

You must log in to post.