Simulation-to-real generalization for deep-learning-based refraction-corrected ultrasound tomography image reconstruction
Physics in Medicine & Biology · 2023
DOI: 10.1088/1361-6560/acaeed ↗Kurz gefasst
Fachartikel in der Zeitschrift Physics in Medicine & Biology, 2023. Verfasst von Wenzhao Zhao und 6 weiteren Autor:innen. Beteiligte Einrichtung: Interdisciplinary Center for Scientific Computing (IWR), Central Institute for Computer Engineering (ZITI), Mannheim Institute for Intelligent System…. Hauptthemen laut MEDLINE: Deep Learning. Die Arbeit wurde bisher 7-mal zitiert (Crossref) und nennt 65 Literaturangaben. Gefördert von: Deutsche Forschungsgemeinschaft.
Abstract (Originalsprache)
Objective . The image reconstruction of ultrasound computed tomography is computationally expensive with conventional iterative methods. The fully learned direct deep learning reconstruction is promising to speed up image reconstruction significantly. However, for direct reconstruction from measurement data, due to the lack of real labeled data, the neural network is usually trained on a simulation dataset and shows poor performance on real data because of the simulation-to-real gap. Approach . To improve the simulation-to-real generalization of neural networks, a series of strategies are developed including a Fourier-transform-integrated neural network, measurement-domain data augmentation methods, and a self-supervised-learning-based patch-wise preprocessing neural network. Our strategies are evaluated on both the simulation dataset and real measurement datasets from two different prototype machines. Main results . The experimental results show that our deep learning methods help to improve the neural networks’ robustness against noise and the generalizability to real measurement data. Significance . Our methods prove that it is possible for neural networks to achieve superior performance to traditional iterative reconstruction algorithms in imaging quality and allow for real-time 2D-image reconstruction. This study helps pave the path for the application of deep learning methods to practical ultrasound tomography image reconstruction based on simulation datasets.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Interdisciplinary Center for Scientific Computing (IWR), Central Institute for Computer Engineering (ZITI), Mannheim Institute for Intelligent Systems in Medicine (MIISM), Medical Faculty Mannheim, Heidelberg University…
- School of Computer Science and Technology, Donghua University, 2999 North Renmin Road, 201620 Shanghai, People's Republic of China.
- Institute for Data Processing and Electronics, Karlsruhe Institute of Technology (KIT), Campus Nord, P.O. Box 3640, D-76021 Karlsruhe, Germany.
- Department of Imaging Physics, Delft University of Technology, Delft, The Netherlands.
- Interdisciplinary Center for Scientific Computing (IWR), Central Institute for Computer Engineering (ZITI), CZS Heidelberg Center for Model-Based AI, Mannheim Institute for Intelligent Systems in Medicine (MIISM), Medic…
ORCID-Profile: Wenzhao Zhao, Koen W A van Dongen, Jürgen Hesser
Förderung
- Deutsche Forschungsgemeinschaft
Zitieren
APA
Zhao, W., Fan, Y., Wang, H., Gemmeke, H., van Dongen, K. W. A., Hopp, T., & Hesser, J. (2023). Simulation-to-real generalization for deep-learning-based refraction-corrected ultrasound tomography image reconstruction. Physics in Medicine & Biology. https://doi.org/10.1088/1361-6560/acaeed
Vancouver
Zhao W, Fan Y, Wang H, Gemmeke H, van Dongen KWA, Hopp T, et al. Simulation-to-real generalization for deep-learning-based refraction-corrected ultrasound tomography image reconstruction. Physics in Medicine & Biology. 2023. doi:10.1088/1361-6560/acaeed
BibTeX
@article{Zhao2023,
author = {Zhao, Wenzhao and Fan, Yuling and Wang, Hongjian and Gemmeke, Hartmut and van Dongen, Koen W A and Hopp, Torsten and Hesser, Jürgen},
title = {Simulation-to-real generalization for deep-learning-based refraction-corrected ultrasound tomography image reconstruction},
journal = {Physics in Medicine & Biology},
year = {2023},
doi = {10.1088/1361-6560/acaeed}
}
EUVIMED