k-Wave

1. Hussain
   

   Member
   Joined: Feb '12
   Posts: 14

   Hello Bradley and Ben,

   I am indeed thankful to you for the latest version release. I'm using the focused detector example for my simulations. I have couple of questions

   1. In most of the examples discussed the source is either disk or sphere; My source depicts the model of tissue or blood vessel so is it possible to use a cubical object as source?

   2. I need to use a spherical focused transducer. I used program similar to the example "focused detector in 3D", the issue is i need to design a transducer of 1MHz frequency and the diameter of my transducer is lesser than the focal length. So after the simulations are finished my objects looks very close to the detector. I saw the example under the section ultrasound simulations, but the transducer is a linear one. Is there a possibility to design the transducer. Also how can i define the frequency of the transducer?

   Thanks a lot for your time to answer to the questions

   Regards,

   Hussain

   Posted 12 years ago #

2. Hussain
   

   Member
   Joined: Feb '12
   Posts: 14

   Pls do find the attached image of my detector and object

   

   Posted 12 years ago #

3. Bradley Treeby
   

   Developer
   Joined: Mar '10
   Posts: 1,640

   Hi Hussain,

   1. Quick question: Are you trying to model a photoacoustic source (initial value problem) or a time varying source? You can model arbitrary photoacoustic sources simply by assigning your source distribution to source.p0 (which is a 3D matrix the same size as your grid). For time varying sources, you can similarly assign source.p_mask to be any distribution. You can then supply either one time series which is applied to all source points, or a separate one for each source point.

   2. Do you have dimensions for the height or width of your transducer in addition to the radius? This will also affect the effective focal length of the transducer you have created. To see what I mean in the case where the transducer is used a source, open up the m-file for the Simulating Transducer Field Patterns Example and replace line 50 (the definition of the source mask) with

   height = 5;
source.p_mask = makeCircle(Nx, Ny, 61, Ny/2, 60, pi);
source.p_mask(height:end, :) = 0;

   Try changing the height value to see what happens.

   To simulate a particular frequency response, you will need to filter the pressure values returned by k-Wave. For example, you could try using gaussianFilter.

   As there seem to be a number of users interested in curved ultrasound transducers, we will try and include some additional functions and examples in the next k-Wave release.

   I hope that helps,

   Brad.

   Posted 12 years ago #

4. Hussain
   

   Member
   Joined: Feb '12
   Posts: 14

   Dear Brad,

   Thanks a lot for your prompt reply.

   1. My source is based on initial value problem. Right now i'm using makeball function to create the source. I need to model the source replicating a tissue with dimensions 5*5*5cm. I have the value for initial pressure distribution which is the product of fluence and absorption co-efficient. How can i use the value for modelling in k-wave?

   2. As you pointed out, i tried using the height of the transducer and it's now working.. Thanks a lot!!!

   3. The center frequency of my detector transducer is say 1MHz. How do i use the gaussianFilter function to design the transducer. I would also like to see, the variations in the signal when the transducer frequency changes?

   Regards,

   Hussain

   Posted 12 years ago #

5. Bradley Treeby
   

   Developer
   Joined: Mar '10
   Posts: 1,640

   Hi Hussain,

   1. Do you mean the entire domain of interest is 5 x 5 x 5 cm, or there is a small optically absorbing sub-region that you are interested in?

   In 3D, source.p0 is simply a 3D matrix which describes the initial pressure (in units of Pa) at each grid point within the domain. You can create this matrix and assign the grid point values (i.e., matrix elements) in the same way as you would create any other matrix in MATLAB. In your case, as you say, you can obtain this distribution from product of the fluence, absorption coefficient, and Gruneisen parameter at each grid point.

   The size of certain objects within the domain (as represented by a certain number of grid points or matrix elements) can be calculated using the values for kgrid.dx, kgrid.dy, and kgrid.dz.

   2. The gaussianFilter function is fairly straightforward to use. Simply choose a value for the center frequency and bandwidth (as a percentage of the centre frequency). For example:

   sensor_data_filtered = gaussianFilter(sensor_data, 1/kgrid.dt, center_freq, bandwidth);

   You can then apply this filter for different values of center_freq (in units of Hz) to see the effect on your signals.

   Brad.

   Posted 12 years ago #

RSS feed for this topic

Reply

You must log in to post.