Challenge on Ultrasound Beamforming with Deep Learning (CUBDL)

Muyinatu A. Lediju Bell; Jiaqi Huang; Dongwoon Hyun; Yonina C. Eldar; Ruud van Sloun; Massimo Mischi

doi:10.1109/IUS46767.2020.9251434

Challenge on Ultrasound Beamforming with Deep Learning (CUBDL)

Muyinatu A. Lediju Bell
, Jiaqi Huang
, Dongwoon Hyun
, Yonina C. Eldar
, Ruud van Sloun
, Massimo Mischi

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

51 Citations (Scopus)

1 Downloads (Pure)

Abstract

Recent developments in deep learning have created immense potential for improving ultrasound beamforming. We organized a Challenge on Ultrasound Beamforming with Deep Learning (CUBDL) to benchmark methodologies in this space with two transmission data types: plane wave and focused transmissions. Plane wave ultrasound transmissions have created new opportunities for ultrafast ultrasound imaging, while focused ultrasound transmissions are more traditional and are widely used in most clinical ultrasound systems available today. For both transmission types, we challenged participants to obtain the best image quality under the fastest possible frame rates. CUBDL organizers solicited datasets from several leading ultrasound groups around the world and received a total of 106 data sequences including in vivo, ex vivo, simulated, and experimental phantom datasets. These submissions formed our test datasets, which were not released to participants while the challenge was open. The challenge was composed of three optional tasks (one including two subtasks) that were evaluated using the test datasets. Participants had the option to provide their results for a minimum of one up to a maximum of four tasks or subtasks: (1) beamforming with deep learning after a single plane wave transmission, which had two subtasks to either (a) match or (b) exceed traditional image quality metrics obtained with multiple plane wave transmissions; (2) beamforming with deep learning after a few plane wave transmissions; (3) beamforming with deep learning to achieve dynamic transmit focusing from datasets acquired with a single transmit focus. Evaluation included image quality metrics as well as network complexity metrics. A challenge website was created to provide information and updates: https://cubdl.jhu.edu/.

Original language	English
Title of host publication	2020 IEEE International Ultrasonics Symposium (IUS)
Publisher	Institute of Electrical and Electronics Engineers
Number of pages	5
ISBN (Electronic)	978-1-7281-5448-0
DOIs	https://doi.org/10.1109/IUS46767.2020.9251434
Publication status	Published - 17 Nov 2020
Event	2020 IEEE International Ultrasonics Symposium, IUS 2020 - Las Vegas, United States Duration: 7 Sept 2020 → 11 Sept 2020

Conference

Conference	2020 IEEE International Ultrasonics Symposium, IUS 2020
Country/Territory	United States
City	Las Vegas
Period	7/09/20 → 11/09/20

Bibliographical note

Funding Information:
We thank our sponsors (Verasonics, Inc., IEEE, and IUS) for their generous support of this challenge, as well as the many individuals and groups who contributed data to make this challenge possible. The following individuals provided data from their institutions: Alycen Wiacek, Jiaqi Huang, and Muyinatu Bell from Johns Hopkins University; Ping Gong and Shigao Chen from Mayo Clinic; Ben Luijten and Massimo Mischi from Eindhoven University of Technology; Xi Zhang and Jiawen Luo from Tsinghua University; Magnus Cinthio from Lund University; Vincent Perrot and Hervé Liebgott from Insa Lyon; Ole Marius Hoel Rindal from University of Oslo; Alessandro Ramalli and Piero Tortoli from University of Florence; Julien Grondin and Elisa Konofagou from Columbia University. We also thank Olivier Bernard for creating data sequence OSL010 described in the CUBDL Data Guide available on the challenge website [10] and Hervé Liebgott for his helpful advice. M.A.L.B. additionally acknowledges NIH Trailblazer Award R21 EB025621 for partial support of her time on this project.

Funding

We thank our sponsors (Verasonics, Inc., IEEE, and IUS) for their generous support of this challenge, as well as the many individuals and groups who contributed data to make this challenge possible. The following individuals provided data from their institutions: Alycen Wiacek, Jiaqi Huang, and Muyinatu Bell from Johns Hopkins University; Ping Gong and Shigao Chen from Mayo Clinic; Ben Luijten and Massimo Mischi from Eindhoven University of Technology; Xi Zhang and Jiawen Luo from Tsinghua University; Magnus Cinthio from Lund University; Vincent Perrot and Hervé Liebgott from Insa Lyon; Ole Marius Hoel Rindal from University of Oslo; Alessandro Ramalli and Piero Tortoli from University of Florence; Julien Grondin and Elisa Konofagou from Columbia University. We also thank Olivier Bernard for creating data sequence OSL010 described in the CUBDL Data Guide available on the challenge website [10] and Hervé Liebgott for his helpful advice. M.A.L.B. additionally acknowledges NIH Trailblazer Award R21 EB025621 for partial support of her time on this project.

Access to Document

10.1109/IUS46767.2020.9251434

Cite this

@inproceedings{13533763c83447c8a528c48f9b9e124f,

title = "Challenge on Ultrasound Beamforming with Deep Learning (CUBDL)",

abstract = "Recent developments in deep learning have created immense potential for improving ultrasound beamforming. We organized a Challenge on Ultrasound Beamforming with Deep Learning (CUBDL) to benchmark methodologies in this space with two transmission data types: plane wave and focused transmissions. Plane wave ultrasound transmissions have created new opportunities for ultrafast ultrasound imaging, while focused ultrasound transmissions are more traditional and are widely used in most clinical ultrasound systems available today. For both transmission types, we challenged participants to obtain the best image quality under the fastest possible frame rates. CUBDL organizers solicited datasets from several leading ultrasound groups around the world and received a total of 106 data sequences including in vivo, ex vivo, simulated, and experimental phantom datasets. These submissions formed our test datasets, which were not released to participants while the challenge was open. The challenge was composed of three optional tasks (one including two subtasks) that were evaluated using the test datasets. Participants had the option to provide their results for a minimum of one up to a maximum of four tasks or subtasks: (1) beamforming with deep learning after a single plane wave transmission, which had two subtasks to either (a) match or (b) exceed traditional image quality metrics obtained with multiple plane wave transmissions; (2) beamforming with deep learning after a few plane wave transmissions; (3) beamforming with deep learning to achieve dynamic transmit focusing from datasets acquired with a single transmit focus. Evaluation included image quality metrics as well as network complexity metrics. A challenge website was created to provide information and updates: https://cubdl.jhu.edu/.",

author = "\{Lediju Bell\}, \{Muyinatu A.\} and Jiaqi Huang and Dongwoon Hyun and Eldar, \{Yonina C.\} and \{van Sloun\}, Ruud and Massimo Mischi",

note = "Funding Information: We thank our sponsors (Verasonics, Inc., IEEE, and IUS) for their generous support of this challenge, as well as the many individuals and groups who contributed data to make this challenge possible. The following individuals provided data from their institutions: Alycen Wiacek, Jiaqi Huang, and Muyinatu Bell from Johns Hopkins University; Ping Gong and Shigao Chen from Mayo Clinic; Ben Luijten and Massimo Mischi from Eindhoven University of Technology; Xi Zhang and Jiawen Luo from Tsinghua University; Magnus Cinthio from Lund University; Vincent Perrot and Herv{\'e} Liebgott from Insa Lyon; Ole Marius Hoel Rindal from University of Oslo; Alessandro Ramalli and Piero Tortoli from University of Florence; Julien Grondin and Elisa Konofagou from Columbia University. We also thank Olivier Bernard for creating data sequence OSL010 described in the CUBDL Data Guide available on the challenge website [10] and Herv{\'e} Liebgott for his helpful advice. M.A.L.B. additionally acknowledges NIH Trailblazer Award R21 EB025621 for partial support of her time on this project. ; 2020 IEEE International Ultrasonics Symposium, IUS 2020 ; Conference date: 07-09-2020 Through 11-09-2020",

year = "2020",

month = nov,

day = "17",

doi = "10.1109/IUS46767.2020.9251434",

language = "English",

booktitle = "2020 IEEE International Ultrasonics Symposium (IUS)",

publisher = "Institute of Electrical and Electronics Engineers",

address = "United States",

}

Lediju Bell, MA, Huang, J, Hyun, D, Eldar, YC, van Sloun, R & Mischi, M 2020, Challenge on Ultrasound Beamforming with Deep Learning (CUBDL). in 2020 IEEE International Ultrasonics Symposium (IUS)., 9251434, Institute of Electrical and Electronics Engineers, 2020 IEEE International Ultrasonics Symposium, IUS 2020, Las Vegas, Nevada, United States, 7/09/20. https://doi.org/10.1109/IUS46767.2020.9251434

TY - GEN

T1 - Challenge on Ultrasound Beamforming with Deep Learning (CUBDL)

AU - Lediju Bell, Muyinatu A.

AU - Huang, Jiaqi

AU - Hyun, Dongwoon

AU - Eldar, Yonina C.

AU - van Sloun, Ruud

AU - Mischi, Massimo

N1 - Funding Information: We thank our sponsors (Verasonics, Inc., IEEE, and IUS) for their generous support of this challenge, as well as the many individuals and groups who contributed data to make this challenge possible. The following individuals provided data from their institutions: Alycen Wiacek, Jiaqi Huang, and Muyinatu Bell from Johns Hopkins University; Ping Gong and Shigao Chen from Mayo Clinic; Ben Luijten and Massimo Mischi from Eindhoven University of Technology; Xi Zhang and Jiawen Luo from Tsinghua University; Magnus Cinthio from Lund University; Vincent Perrot and Hervé Liebgott from Insa Lyon; Ole Marius Hoel Rindal from University of Oslo; Alessandro Ramalli and Piero Tortoli from University of Florence; Julien Grondin and Elisa Konofagou from Columbia University. We also thank Olivier Bernard for creating data sequence OSL010 described in the CUBDL Data Guide available on the challenge website [10] and Hervé Liebgott for his helpful advice. M.A.L.B. additionally acknowledges NIH Trailblazer Award R21 EB025621 for partial support of her time on this project.

PY - 2020/11/17

Y1 - 2020/11/17

N2 - Recent developments in deep learning have created immense potential for improving ultrasound beamforming. We organized a Challenge on Ultrasound Beamforming with Deep Learning (CUBDL) to benchmark methodologies in this space with two transmission data types: plane wave and focused transmissions. Plane wave ultrasound transmissions have created new opportunities for ultrafast ultrasound imaging, while focused ultrasound transmissions are more traditional and are widely used in most clinical ultrasound systems available today. For both transmission types, we challenged participants to obtain the best image quality under the fastest possible frame rates. CUBDL organizers solicited datasets from several leading ultrasound groups around the world and received a total of 106 data sequences including in vivo, ex vivo, simulated, and experimental phantom datasets. These submissions formed our test datasets, which were not released to participants while the challenge was open. The challenge was composed of three optional tasks (one including two subtasks) that were evaluated using the test datasets. Participants had the option to provide their results for a minimum of one up to a maximum of four tasks or subtasks: (1) beamforming with deep learning after a single plane wave transmission, which had two subtasks to either (a) match or (b) exceed traditional image quality metrics obtained with multiple plane wave transmissions; (2) beamforming with deep learning after a few plane wave transmissions; (3) beamforming with deep learning to achieve dynamic transmit focusing from datasets acquired with a single transmit focus. Evaluation included image quality metrics as well as network complexity metrics. A challenge website was created to provide information and updates: https://cubdl.jhu.edu/.

AB - Recent developments in deep learning have created immense potential for improving ultrasound beamforming. We organized a Challenge on Ultrasound Beamforming with Deep Learning (CUBDL) to benchmark methodologies in this space with two transmission data types: plane wave and focused transmissions. Plane wave ultrasound transmissions have created new opportunities for ultrafast ultrasound imaging, while focused ultrasound transmissions are more traditional and are widely used in most clinical ultrasound systems available today. For both transmission types, we challenged participants to obtain the best image quality under the fastest possible frame rates. CUBDL organizers solicited datasets from several leading ultrasound groups around the world and received a total of 106 data sequences including in vivo, ex vivo, simulated, and experimental phantom datasets. These submissions formed our test datasets, which were not released to participants while the challenge was open. The challenge was composed of three optional tasks (one including two subtasks) that were evaluated using the test datasets. Participants had the option to provide their results for a minimum of one up to a maximum of four tasks or subtasks: (1) beamforming with deep learning after a single plane wave transmission, which had two subtasks to either (a) match or (b) exceed traditional image quality metrics obtained with multiple plane wave transmissions; (2) beamforming with deep learning after a few plane wave transmissions; (3) beamforming with deep learning to achieve dynamic transmit focusing from datasets acquired with a single transmit focus. Evaluation included image quality metrics as well as network complexity metrics. A challenge website was created to provide information and updates: https://cubdl.jhu.edu/.

UR - https://www.scopus.com/pages/publications/85097886749

U2 - 10.1109/IUS46767.2020.9251434

DO - 10.1109/IUS46767.2020.9251434

M3 - Conference contribution

AN - SCOPUS:85097886749

BT - 2020 IEEE International Ultrasonics Symposium (IUS)

PB - Institute of Electrical and Electronics Engineers

T2 - 2020 IEEE International Ultrasonics Symposium, IUS 2020

Y2 - 7 September 2020 through 11 September 2020

ER -

Challenge on Ultrasound Beamforming with Deep Learning (CUBDL)

Abstract

Conference

Bibliographical note

Funding

Access to Document

Other files and links

Fingerprint

Cite this