k-Wave

1. AhmetYildirim
   

   Member
   Joined: Aug '20
   Posts: 3

   I am trying to make photoacoustic image reconstruction of a heterogenous medium that I've created on k-wave.

   The properties of this simulated area are basically:
   I have 3 mediums with different sound speeds, their density is equal to each other. And i have a binary line sensor on top of these mediums.

   My aim is to determine the sound speed of each medium and obtain better reconstruction image. I need to make correlation between sound speed of mediums and their thickness. Until now i used Time Reversal reconstruction algorithm given with k-wave toolbox, i got result image of various combination of sound speeds for each medium but it is hard and time consuming to make correlation between thousands of images.

   There is "Automatic Sound Speed Selection Example" given in k-wave toolbox. I tried to determine sound speeds of the mediums with that but i couldn't achieve any meaningful output (at least for me). I think it is because i have 3 different mediums and a line sensor.

   So, my question is that am i able to determine the sound speeds of mediums that i have created in k-wave by only having sensor_data as an input? If so how can i do that to determine sound speeds of each medium.

   Posted 4 years ago #

2. Bradley Treeby
   

   Developer
   Joined: Mar '10
   Posts: 1,643

   Have you tried using a simple optimisation algorithm to find the three sound speeds by maximising the sharpness metric?

   I recall someone published on this a few years ago, but I don't remember the reference off hand. You could try going through the citations of the original autofocus paper on google scholar to see if you can find it.

   Brad

   Posted 4 years ago #

3. AhmetYildirim
   

   Member
   Joined: Aug '20
   Posts: 3

   Thank you for your answer. Yes i have tried the algorithm which maximizes the sharpness metric. But the plot i obtained has several peaks close to each other. Also the Normalized Focus Function value of the real sound speed combination is not the one with highest value.

   When i give the correct sound speeds of 2 of 3 mediums and change the other one through loop and calculate the sharpness, it mostly gives my the correct sound speed of that medium. There is no problem with that.

   But when i loop all mediums sound speeds and compare sharpness values of all combinations it has a lot close peaks as i mentioned.

   What could be the reason?

   As an example:
   Real values of sound speeds of mediums are: 1450, 1550, 1600
   I created a nested loop and changed sound speeds as 1400:25:1700
   The result has peak at almost 15 combination and the highest one is at 1450, 1550, 1400

   I suppose it is because of the medium with really low thickness compared to the others. Which is the topmost medium. Sharpness decreases when i increase the sound speed assumption of that medium.

   Posted 4 years ago #

4. Bradley Treeby
   

   Developer
   Joined: Mar '10
   Posts: 1,643

   Hi Ahmet,

   This suggests that the inverse problem as you've set it up is not well-posed. It may be that you need more complexity in your source, or perhaps more sensor data.

   Brad.

   Posted 4 years ago #

RSS feed for this topic

Reply

You must log in to post.