Sensitivity of a partially learned model-based reconstruction algorithm

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademic

2 Downloads (Pure)

Abstract

We replace part of a model-based iterative algorithm with a convolutional neural network in order to improve the quality of tomography reconstructions. We analyse its robustness against uncertainties in the image and uncertainties in system settings. Results are presented for the application of photoacoustic tomography in a limited angle setup.
Original languageEnglish
Title of host publication89th Annual Meeting of the International Association of Applied Mathematics and Mechanics (GAMM)
Number of pages2
Volume18
Edition1
DOIs
Publication statusPublished - 17 Dec 2018

Publication series

NameProceedings in Applied Mathematics and Mechanics
PublisherWiley-VCH Verlag
ISSN (Print)1617-7061

Fingerprint

tomography
sensitivity

Keywords

  • inverse problems
  • reconstruction
  • deep learning
  • algorithm

Cite this

Boink, Y. E., Van Gils, S. A., Manohar, S., & Brune, C. (2018). Sensitivity of a partially learned model-based reconstruction algorithm. In 89th Annual Meeting of the International Association of Applied Mathematics and Mechanics (GAMM) (1 ed., Vol. 18). (Proceedings in Applied Mathematics and Mechanics). https://doi.org/10.1002/pamm.201800222
Boink, Yoeri E. ; Van Gils, Stephan A. ; Manohar, Srirang ; Brune, Christoph. / Sensitivity of a partially learned model-based reconstruction algorithm. 89th Annual Meeting of the International Association of Applied Mathematics and Mechanics (GAMM). Vol. 18 1. ed. 2018. (Proceedings in Applied Mathematics and Mechanics).
@inproceedings{19494de7dab1456f818c8b33d47e9675,
title = "Sensitivity of a partially learned model-based reconstruction algorithm",
abstract = "We replace part of a model-based iterative algorithm with a convolutional neural network in order to improve the quality of tomography reconstructions. We analyse its robustness against uncertainties in the image and uncertainties in system settings. Results are presented for the application of photoacoustic tomography in a limited angle setup.",
keywords = "inverse problems, reconstruction, deep learning, algorithm",
author = "Boink, {Yoeri E.} and {Van Gils}, {Stephan A.} and Srirang Manohar and Christoph Brune",
year = "2018",
month = "12",
day = "17",
doi = "10.1002/pamm.201800222",
language = "English",
volume = "18",
series = "Proceedings in Applied Mathematics and Mechanics",
publisher = "Wiley-VCH Verlag",
booktitle = "89th Annual Meeting of the International Association of Applied Mathematics and Mechanics (GAMM)",
edition = "1",

}

Boink, YE, Van Gils, SA, Manohar, S & Brune, C 2018, Sensitivity of a partially learned model-based reconstruction algorithm. in 89th Annual Meeting of the International Association of Applied Mathematics and Mechanics (GAMM). 1 edn, vol. 18, Proceedings in Applied Mathematics and Mechanics. https://doi.org/10.1002/pamm.201800222

Sensitivity of a partially learned model-based reconstruction algorithm. / Boink, Yoeri E.; Van Gils, Stephan A.; Manohar, Srirang; Brune, Christoph.

89th Annual Meeting of the International Association of Applied Mathematics and Mechanics (GAMM). Vol. 18 1. ed. 2018. (Proceedings in Applied Mathematics and Mechanics).

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademic

TY - GEN

T1 - Sensitivity of a partially learned model-based reconstruction algorithm

AU - Boink, Yoeri E.

AU - Van Gils, Stephan A.

AU - Manohar, Srirang

AU - Brune, Christoph

PY - 2018/12/17

Y1 - 2018/12/17

N2 - We replace part of a model-based iterative algorithm with a convolutional neural network in order to improve the quality of tomography reconstructions. We analyse its robustness against uncertainties in the image and uncertainties in system settings. Results are presented for the application of photoacoustic tomography in a limited angle setup.

AB - We replace part of a model-based iterative algorithm with a convolutional neural network in order to improve the quality of tomography reconstructions. We analyse its robustness against uncertainties in the image and uncertainties in system settings. Results are presented for the application of photoacoustic tomography in a limited angle setup.

KW - inverse problems

KW - reconstruction

KW - deep learning

KW - algorithm

U2 - 10.1002/pamm.201800222

DO - 10.1002/pamm.201800222

M3 - Conference contribution

VL - 18

T3 - Proceedings in Applied Mathematics and Mechanics

BT - 89th Annual Meeting of the International Association of Applied Mathematics and Mechanics (GAMM)

ER -

Boink YE, Van Gils SA, Manohar S, Brune C. Sensitivity of a partially learned model-based reconstruction algorithm. In 89th Annual Meeting of the International Association of Applied Mathematics and Mechanics (GAMM). 1 ed. Vol. 18. 2018. (Proceedings in Applied Mathematics and Mechanics). https://doi.org/10.1002/pamm.201800222