Authorization

 

Journal

Volume 23, №1

 
Encoder neural network in 2D acoustic tomography (pp.83-98)
A.Yu. Prikhodko1, M.A. Shishlenin1,2,3, N.S. Novikov1,2,3, D.V. Klyuchinskiy2
https://doi.org/10.30546/1683-6154.23.1.2024.83


1Novosibirsk State University, Pirogova St., 2, Novosibirsk, Russia, e-mail: a.prikhodko@g.nsu.ru

2Institute of Computational Mathematics and Mathematical Geophysics, Acad. Lavrentiev Ave., 6, Novosibirsk, Russia, e-mail: m.a.shishlenin@mail.ru

3Sobolev Institute of Mathematics, Akad. Koptyuga avenue, 4, 630090, Novosibirsk, Russia

∗Corresponding author e-mail: novikov-1989@yandex.ru

 

Abstract. We investigate deep learning approach in 2D dynamic ultrasound acoustic tomography. The mathematical model of acoustic tomography is described by a first-order hyperbolic system PDE and is based on conservation laws. This model guarantees us that the training sets of dynamic data are close to the physical solution. We train a neural network consisting of an encoder and a decoder with this data (they contain only one inclusion) and associate the data with a velocity coefficient. Numerical results show that we recover not only single inclusions, but also homogeneities consisting of two inclusions.

 

Keywords: Acoustic Tomography, Ultrasound, Deep Learning, Neural Networks, Coefficient Inverse Problem.

 

AMS Subject Classification: 35R30, 68T07.



COIA 2027

7-9 July 2027
Baku, Azerbaijan

Search