Hello all,
I am trying to use the focused detector example in 3D simulations. The example describes the focused bowl detector. I would like to design a detector with a radius of 12 mm and the focal length of 38mm. Is there a way i could use to design a 3D detector with that specification?
In the same example, the source is a time varying sinusoidal one. What happens if i use a time independent source and is it possible to use one?
Thanks a lot for the support and prompt responses.
Hi ranzakhuss,
If the detector is geometrically focused, you can create it using makeSphere. To make it a given size, convert between grid points and meters using the grid spacing (e.g., kgrid.dx). After you've made the complete sphere, mask out the part you don't need following the example.
It's possible to implement a wide range of different source conditions. Take a look at the other examples to see how this works.
Good luck with your simulations,
Brad.
Hello Bradley,
I am confused with the explanation. I would be really helpful if you give me an example.
I had created the computational k-grid with Nx,Ny,Nz= 100 and dx,dy,dz=0.1e-3 (10*10*10).
My radius is 12mm and focal length is 38mm. Initially i used sphere = makeSphere(sphere_sz,sphere_sz,sphere_sz,sphere_radius) in order to make the sphere.
Thanks
Hi ranzakhuss,
The way to model your detector will depend on how it is physically constructed. If the face of your transducer is curved (and thus the focusing comes from the transducer geometry), you can model this by creating a geometrically accurate model of the transducer face. In the case of a spherically focused transducer, you can use makeSphere. The sphere radius will be equal to your geometric focal length (to convert from meters to grid points, divide by the grid point spacing). After creating the sphere, mask out the parts you don't need (by setting them to zero) using the physical dimensions of your transducer, for example the height or width.
If your focusing instead comes from a lens attached to the outer surface of the transducer, you can instead create a flat transducer and model the focusing by applying delays to the recorded signals before summing them. You can define the delays geometrically based on the calculated distance between the focal point and each point on your transducer (i.e., each point in your sensor mask). You can convert from distance to time based on the sound speed in the medium, and from time to time points by dividing by dt.
I hope that helps. If you're still stuck, I'd recommend getting a good book on ultrasound, for example, "Diagnostic Ultrasound Imaging: Inside Out" by Thomas Szabo.
Kind regards,
Brad.
Hello Brad and Ben,
1. In the example- focused detector in 3D. The sensor/detector is created by using MakeSphere function. I am not able to understand how the value for sphere_sz is given. In the previous posts, you had mentioned that the radius of the sphere will be the focal length of the transducer. so i used the value of focal length for the radius of the sphere. But i am not able to find the correct value for Sphere_sz.
2. I am also not able to understand this line
sensor.mask(Nx/2:Nx/2+sphere_sz-1, Ny/2:Ny/2+sphere_sz-1, sphere_offset:sphere_offset+sphere_sz-1) = sphere; How exactly we create the sensor mask?
3. Say i have a computational grid with number of grid points along the x,y and z be 100 and grid spacing is 1e-3. I have created the initial pressure distribution for the source using makeBall function and the radius of my source is 2mm. I would like to fix the position of my detector and see the output by moving the objects along different positions. meaning i can see the output only when the source is in focus with the detector. Is there a way to do this set-up?
Kindly apologize for my inappropriate questions. I have been working hard on these simulations but getting struck with these simple stuffs..
Regards.
Hi ranzakhuss,
Sorry for the delay. I will try and answer your questions in turn.
1. If you open the k-Wave documentation (click on the blue question mark icon from the main MATLAB window and then select "k-Wave Toolbox" from the contents page), you can find detailed documentation for all of the k-Wave functions. If you select makeSphere from the alphabetical list of functions, you can see what the different function inputs mean. You can also see this information from the command line by typing
>> help makeSphere.
In this case, the first three inputs are the x, y, z size of the grid in units of grid points (or voxels). Try experimenting with makeSphere using the same radius but different values for the grid size to see exactly what this does. You can plot the results using voxelPlot.
2. In this particular example, the matrix that represents the transducer (assigned to the variable sphere) is smaller than the overall grid size. The line you mention simply places the transducer into an empty sensor mask that has dimensions Nx by Ny by Nz (it's a binary sensor mask in this case).
3. You can achieve this by moving the position of the source (assigned to source.mask) and running multiple simulations. Take a look at the Modelling Sensor Directivity in 3D Example. If you mean dynamically moving the source within a single simulation, take a look at The Doppler Effect Example.
If you are going to be working with k-Wave a lot, it's probably a good idea to spend some time with the Initial Value Problem examples (even if this is not exactly what you want to simulate). These will introduce you to a lot of the general functionality of the toolbox.
Kind regards,
Brad.
hello ranzakhuss,
Today I see the post you posted six years ago about "focused detector" ,now I have the same question like you.Did you figure it out? Can you help me to design a 3D PTF transducer with a radius of 12 mm and the focal length of 50mm?
Thank you in advance!
Hi peiliu,
There is a new function in k-Wave to easily create a focused detector called makeBowl. This should do what you want.
Brad.
You must log in to post.