Deep learning for fast adaptive beamforming

Ben Luijten, Regev Cohen, Frederik J. de Bruijn, Harold A.W. Schmeitz, Massimo Mischi, Yonina C. Eldar, Ruud J.G. van Sloun

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

2 Downloads (Pure)

Abstract

The real-time nature that makes diagnostic ultrasonography so appealing to clinicians imposes strong constraints on the computational complexity of image reconstruction algorithms. As such, these typically rely on traditional delay-and-sum beamforming, a low-complexity approach that unfortunately comes at the cost of reduced image quality as compared to more advanced and content-adaptive beamformers. Here, we propose a model-aware deep learning strategy to ultrasound image reconstruction, which leverages knowledge of minimum variance beamforming while exploiting the efficiency of deep neural networks. Our approach yields high quality images with strong contrast at real-time reconstruction rates. The neural network is trained using in vivo and simulated radio frequency channel data of a single plane wave transmit, and corresponding high-quality minimum-variance beamformed reconstructions. Performance is benchmarked using simulated acquisitions from the PICMUS [1] dataset, demonstrating the convincing generalizability and image quality of the proposed beamformer.

Original languageEnglish
Title of host publication2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings
Place of PublicationPiscataway
PublisherInstitute of Electrical and Electronics Engineers
Pages1333-1337
Number of pages5
ISBN (Electronic)978-1-4799-8131-1
DOIs
Publication statusPublished - 1 May 2019
Event44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Brighton, United Kingdom
Duration: 12 May 201917 May 2019

Conference

Conference44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019
CountryUnited Kingdom
CityBrighton
Period12/05/1917/05/19

Fingerprint

Beamforming
Image quality
Image reconstruction
Ultrasonography
Computational complexity
Ultrasonics
Neural networks
Deep learning

Keywords

  • Adaptive Beamforming
  • Deep Learning
  • Plane Wave Imaging
  • Ultrasound

Cite this

Luijten, B., Cohen, R., de Bruijn, F. J., Schmeitz, H. A. W., Mischi, M., Eldar, Y. C., & van Sloun, R. J. G. (2019). Deep learning for fast adaptive beamforming. In 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings (pp. 1333-1337). [8683478] Piscataway: Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/ICASSP.2019.8683478
Luijten, Ben ; Cohen, Regev ; de Bruijn, Frederik J. ; Schmeitz, Harold A.W. ; Mischi, Massimo ; Eldar, Yonina C. ; van Sloun, Ruud J.G. / Deep learning for fast adaptive beamforming. 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings. Piscataway : Institute of Electrical and Electronics Engineers, 2019. pp. 1333-1337
@inproceedings{05121f9869a2452aaef022ebd3bff304,
title = "Deep learning for fast adaptive beamforming",
abstract = "The real-time nature that makes diagnostic ultrasonography so appealing to clinicians imposes strong constraints on the computational complexity of image reconstruction algorithms. As such, these typically rely on traditional delay-and-sum beamforming, a low-complexity approach that unfortunately comes at the cost of reduced image quality as compared to more advanced and content-adaptive beamformers. Here, we propose a model-aware deep learning strategy to ultrasound image reconstruction, which leverages knowledge of minimum variance beamforming while exploiting the efficiency of deep neural networks. Our approach yields high quality images with strong contrast at real-time reconstruction rates. The neural network is trained using in vivo and simulated radio frequency channel data of a single plane wave transmit, and corresponding high-quality minimum-variance beamformed reconstructions. Performance is benchmarked using simulated acquisitions from the PICMUS [1] dataset, demonstrating the convincing generalizability and image quality of the proposed beamformer.",
keywords = "Adaptive Beamforming, Deep Learning, Plane Wave Imaging, Ultrasound",
author = "Ben Luijten and Regev Cohen and {de Bruijn}, {Frederik J.} and Schmeitz, {Harold A.W.} and Massimo Mischi and Eldar, {Yonina C.} and {van Sloun}, {Ruud J.G.}",
year = "2019",
month = "5",
day = "1",
doi = "10.1109/ICASSP.2019.8683478",
language = "English",
pages = "1333--1337",
booktitle = "2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings",
publisher = "Institute of Electrical and Electronics Engineers",
address = "United States",

}

Luijten, B, Cohen, R, de Bruijn, FJ, Schmeitz, HAW, Mischi, M, Eldar, YC & van Sloun, RJG 2019, Deep learning for fast adaptive beamforming. in 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings., 8683478, Institute of Electrical and Electronics Engineers, Piscataway, pp. 1333-1337, 44th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019, Brighton, United Kingdom, 12/05/19. https://doi.org/10.1109/ICASSP.2019.8683478

Deep learning for fast adaptive beamforming. / Luijten, Ben; Cohen, Regev; de Bruijn, Frederik J.; Schmeitz, Harold A.W.; Mischi, Massimo; Eldar, Yonina C.; van Sloun, Ruud J.G.

2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings. Piscataway : Institute of Electrical and Electronics Engineers, 2019. p. 1333-1337 8683478.

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

TY - GEN

T1 - Deep learning for fast adaptive beamforming

AU - Luijten, Ben

AU - Cohen, Regev

AU - de Bruijn, Frederik J.

AU - Schmeitz, Harold A.W.

AU - Mischi, Massimo

AU - Eldar, Yonina C.

AU - van Sloun, Ruud J.G.

PY - 2019/5/1

Y1 - 2019/5/1

N2 - The real-time nature that makes diagnostic ultrasonography so appealing to clinicians imposes strong constraints on the computational complexity of image reconstruction algorithms. As such, these typically rely on traditional delay-and-sum beamforming, a low-complexity approach that unfortunately comes at the cost of reduced image quality as compared to more advanced and content-adaptive beamformers. Here, we propose a model-aware deep learning strategy to ultrasound image reconstruction, which leverages knowledge of minimum variance beamforming while exploiting the efficiency of deep neural networks. Our approach yields high quality images with strong contrast at real-time reconstruction rates. The neural network is trained using in vivo and simulated radio frequency channel data of a single plane wave transmit, and corresponding high-quality minimum-variance beamformed reconstructions. Performance is benchmarked using simulated acquisitions from the PICMUS [1] dataset, demonstrating the convincing generalizability and image quality of the proposed beamformer.

AB - The real-time nature that makes diagnostic ultrasonography so appealing to clinicians imposes strong constraints on the computational complexity of image reconstruction algorithms. As such, these typically rely on traditional delay-and-sum beamforming, a low-complexity approach that unfortunately comes at the cost of reduced image quality as compared to more advanced and content-adaptive beamformers. Here, we propose a model-aware deep learning strategy to ultrasound image reconstruction, which leverages knowledge of minimum variance beamforming while exploiting the efficiency of deep neural networks. Our approach yields high quality images with strong contrast at real-time reconstruction rates. The neural network is trained using in vivo and simulated radio frequency channel data of a single plane wave transmit, and corresponding high-quality minimum-variance beamformed reconstructions. Performance is benchmarked using simulated acquisitions from the PICMUS [1] dataset, demonstrating the convincing generalizability and image quality of the proposed beamformer.

KW - Adaptive Beamforming

KW - Deep Learning

KW - Plane Wave Imaging

KW - Ultrasound

UR - http://www.scopus.com/inward/record.url?scp=85068999220&partnerID=8YFLogxK

U2 - 10.1109/ICASSP.2019.8683478

DO - 10.1109/ICASSP.2019.8683478

M3 - Conference contribution

AN - SCOPUS:85068999220

SP - 1333

EP - 1337

BT - 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings

PB - Institute of Electrical and Electronics Engineers

CY - Piscataway

ER -

Luijten B, Cohen R, de Bruijn FJ, Schmeitz HAW, Mischi M, Eldar YC et al. Deep learning for fast adaptive beamforming. In 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019 - Proceedings. Piscataway: Institute of Electrical and Electronics Engineers. 2019. p. 1333-1337. 8683478 https://doi.org/10.1109/ICASSP.2019.8683478