k-Wave

1. hyeongyu
   

   Member
   Joined: May '23
   Posts: 8

   Hi i'm trying to build acoustic hologram. for mouse neuromodulation. by using k-wave.

   I have a question in two method.

   First, i used time reversal method for building holgram. but in my idea, it's not accurate. And i used equivalent source hologram. in your paper "Equivalent-Source Acoustic Holography for Projecting Measured Ultrasound Fields Through Complex Media" does not provide comparison between time reversal, IASA, equivalent-source methods.

   I think, Equivalent-source method has better accuracy compare with IASA and time reversal. because Equivalent source is better in complex media that including nonlinearity and heterogeneous material properties.

   Is it right? I want to know about your opinion. Thank you for reading despite my poor English.

   Posted 1 year ago #

2. astanziola
   

   Member
   Joined: Oct '20
   Posts: 1

   Hi hyeongyu,

   Agree that the equilavent source method is accurate, as it uses a gradient descent method that aims to maximize the similarity between acoustic fields. Also, to my knowledge only equivalent source and time reversal can deal with heterogeneous material properties, while IASA can't.

   In any case, both equivalent source and time reversal techniques only allow to calculate an ideal source to generate a reference field. They don't directly give a method to translate the source plane into the material properties of an acoustic hologram. So if your aim is to generate an hologram using a physical lens, rather than a multi-element transducer, how this step is implemented will greatly affect the resulting hologram quality.

   For single-frequency linear propagation, we have implemented the analogous of the equivalent-source method that directly optimizes the material properties of the lens here: https://github.com/ucl-bug/holab
   In free space, it has shown improved performance compared to IASA, and it also showed an improvement of the target acoustic field quality for transcranial hologram design (in simulation, see Fig. S3 here https://arxiv.org/abs/2305.03625 ).

   I would expect that the same results would hold for the non-linear case, even if we have not tried that yet.

   Kind regards,
   Antonio

   Posted 1 year ago #

RSS feed for this topic

Reply

You must log in to post.