Deep convolutional neural networks for estimating porous material parameters with ultrasound tomography
| Autor: | Janne M. J. Huttunen, Leo Kärkkäinen, Jan S. Hesthaven, Timo Lähivaara |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
Acoustics and Ultrasonics
Artificial neural network Wave propagation Computer science Estimation theory FOS: Physical sciences Inversion (meteorology) Computational Physics (physics.comp-ph) 010502 geochemistry & geophysics 01 natural sciences Convolutional neural network Tortuosity Ultrasound Tomography Arts and Humanities (miscellaneous) Discontinuous Galerkin method 0103 physical sciences Physics - Computational Physics 010301 acoustics Algorithm 0105 earth and related environmental sciences |
| Zdroj: | The Journal of the Acoustical Society of America. 143:1148-1158 |
| ISSN: | 0001-4966 |
| DOI: | 10.1121/1.5024341 |
| Popis: | The feasibility of data based machine learning applied to ultrasound tomography is studied to estimate water-saturated porous material parameters. In this work, the data to train the neural networks is simulated by solving wave propagation in coupled poroviscoelastic-viscoelastic-acoustic media. As the forward model, a high-order discontinuous Galerkin method is considered, while deep convolutional neural networks are used to solve the parameter estimation problem. In the numerical experiment, the material porosity and tortuosity is estimated, while the remaining parameters which are of less interest are successfully marginalized in the neural networks-based inversion. Computational examples confirm the feasibility and accuracy of this approach. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|