Speaker
Description
The focusing of ultrasonic wavefields can be achieved in a variety of ways, for example through the use of time reversal mirrors. Most of these applications rely on the use of the recorded Green’s function and reinject the time-reversed version of this wavefield into the medium of interest. While this approach does allow for an accurate focus of the wavefield, it also generates a secondary propagating wavefield, after focusing has taken place. As a result, a significant amount of energy remains in the medium of interest. In case the medium of interest is the human head, this is undesirable.
An alternative to this approach, called finite time focusing, makes use of specially designed focusing wavefields, that are designed to focus inside the human brain, while minimizing the amount of secondary wavefield energy. These focusing wavefields can be obtained through the use of the Marchenko method, a popular geophysical application. The Marchenko method relies on the separation of waveforms in time and as such, is quite sensitive to the presence of structures that are small compared to the wavelength of the data. Because the human skull is thin, this is a potential pitfall when using the standard Marchenko method.
As an alternative to the use of the Marchenko method, the focusing wavefields can be obtained through the use of inversion. The focusing wavefield can be related to the Green’s function through convolutions with reflection data and transmission data measured by transducers outside the human head. The retrieval of the focusing function can then be solved as a linear inverse problem.
We demonstrate how the focusing wavefields required for finite time focusing can be retrieved using 2D in-silico data that were numerically generated in a slice of the MIDA skull model. We make use of transducers that are located outside the human head to generate the reflection data and the transmission data. The Green’s functions are simulated using transducers at the same positions. We compare the focusing wavefields that are retrieved using the inversion approach to similar wavefields that are retrieved using the Marchenko method and show the benefits of our approach.
| Preferred Contribution Type | Presentation |
|---|
