Hello Brad and Ben,
I would like to know, how the focused detector works? In the documentation you have mentioned that the directionality arises from spatially averaging across the detector surface. Is there a possibility to see the signals coming only from the inner surface of the bowl detector and not from the outer edges?
Thanks a lot for your efforts to help us all.
Hussain
Mmm, good question.
In answer to the first part, there's nothing clever going on, we just measure the pressure time series at each pixel (or voxel) making up the sensor mask and add them together. What happens is that any sound that comes from the focus of the detector will sum up coherently when the time series' are added, and any sound that originated from outside the focus will not, so will not show up very much in the averaged time series.
At the moment, the sensor mask just picks out the values of the pressure and particle velocity at the sensor points and returns them, so it has no way of telling which direction they came from. I suppose you could put the bowl of detectors on the front of a backing with different material properties (higher density and sound speed) which maybe could be designed to reflect away sound that comes from behind. It would also reflect the sound from the front from one transducer element to another, but it might be an interesting thing to try. Let us know if you get anything worked out with that!
Ben
Hello Ben,
Thanks for your reply. I don't understand what you meant by putting a bowl of detectors on the front of backing with different material properties. I would be happy if you could give me an example for doing this.
I tried the simulations without averaging the signals from the detector, i noticed that the same number of signals even when i object is out of focus. Instead of averaging the signals from the entire surface, Is there a way i could average only selected region of the detector? This is help me analyze from where the signals are reflected?
Is the bowl detector constructed on single layer of transducer elements or has many layers? If it has many layers that could be a reason for multiple reflections.
Thanks,
Hussain
HI Hussain,
I don't understand something. Are you currently seeing reflections? From where? Do you have heterogeneous material properties?
Ben
Hello Ben,
What i was trying to explain to you is. I was running a simulation with an object(in Homogenous medium) on focus and found the pressure vs time signal as shown in the image below. I didn't average the signal over the entire detector surface. I could see multiple signals.
http://imageshack.us/photo/my-images/268/infocusn.jpg/
Now i tried to move the object out of focus and found the following signal.
http://imageshack.us/photo/my-images/834/outoffocusy.jpg
I'm not sure if these signals are reflections or signal coming from the object. Ideally the signal from an out of focus object should be negligible.
I would be happy if you could help me in explaining why i get signal even if the object is out of focus and is there a possibility for me to see the signal that comes only from the inside surface of the detector.
I have one more question also. When i run the simulation, i see a computational screen, which indicates X-Y plane, X-Z plane and Y-Z plane. I' would like to know which axis corresponds to which plane?
Thanks a lot for answering my questions very patiently.
Regards,
Hussain
Hi Hussain,
From the images, it looks as though you are plotting lots of time series all on the same graph. Do you mean to do that? Each of the separate time series corresponds to the pressure measured at a different voxel in the sensor mask, ie. one time series per voxel. If you want to see the output of a bowl detector (rather than an array of single detection points) you need to add these together ie. sum over the relevant dimension.
Hope that helps,
Ben
Hi Ben,
Initially when i was performing the simulations i got the average signal i.e. from the entire detector surface. But somehow i seem to get signal even when the object is far from the detector. I perform the simulation each time by changing the position of the object and my interest is to see the signal only when its in focus. I have some additional questions i'm extremely sorry to trouble you again and again. I am getting few results but not sure if it's the right one, i'm supposed to get.
1. I would like to know if there is a way to find the number of voxels present in the detector?. How each voxel picks up the signal/pressure wave.
2. What do i do to average the signals only from the inner surface of the bowl detector? As you pointed out, if the take the average of the signals i can get the sum of signals from the entire detector. I dont want the average over the entire surface i just need the average from the inner surface.
3. In the Plot_Sim, currently i see the position of my detector alone x-y plane and y-z plane. How to place the detector in order to see the detector along x-z plane?
Thanks for clarifying all the doubts.
Regards,
Hussain
Hi Hussain,
Just to be clear, in k-Wave, the shape and size of the detector has no effect on the acoustic waves, it just defines the grid points in the medium where the time history of the pressure field should be stored. To model a large detector, the sensor is defined as a number of contiguous grid points (in this case using a section of a sphere). As Ben has mentioned, to model a single element focused detector, you need to sum up the time signals stored at each of these grid points so you are left with a single time series. The sensor_data output from the simulation functions is indexed as sensor_data(position, time) so you can sum the signals using something like
signal = sum(sensor_data, 1);
After you have summed the signals, you should see that when the object is in focus, the individual signals will coherently sum to yield a large signal amplitude. When the object is not in focus, the signals will largely cancel each other out, and the overall magnitude will be much smaller. In k-Wave, you will likely always see some signal unless the medium is very large, the absorption values very high, or you add noise to your signals. In practice, the sensitivity of your detector will limit when you are able to still see a signal.
To answer your other questions:
1. You can find the number of grid points in the sensor using sum(sensor.mask(:))
2. I am not sure what you mean by the inner surface of the detector? As the time series are recorded at particular grid points, there is no 'inner' or 'outer' surface.
3. You could either rotate the sensor using permute or alternatively create the detector by masking out part of the sphere in the other directions.
Good luck with your simulations,
Brad.
Hi,
I have downloaded K-Wave toolbox , every application has some bug.iam not able to run even one m file sucessfully. I used correct directoryand defined correct path in matlab.
Hi ravigopal,
Sorry to hear you couldn't get it working. What version of MATLAB are you using and what is the error the you see? If you are in another directory and type help k-Wave does it print the k-Wave help?
Brad.
You must log in to post.
