Model-based Deep Learning on Ultrasound Channel Data for Fast Ultrasound Localization Microscopy

Youn, Jihwan; Luijten, Ben; Schou, Mikkel; Stuart, Matthias Bo; Eldar, Yonina C.; Sloun, Ruud J. G. van; Jensen, Jørgen Arendt

doi:10.1109/ius52206.2021.9593435

Cited by 4 publications

(3 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Youn et al (Youn et al, 2021) took this one step further and combined the image-domain LISTA architecture by with the ABLE beamforming architecture of Luijten et al (Luijten et al, 2020), training the joint network end-toend. The authors found that this outperforms non-joint optimization and that ABLE learns to accommodate the downstream localization problem addressed by LISTA.…”

Section: Post-processing and Interpretationmentioning

confidence: 99%

Ultrasound Signal Processing: From Models to Deep Learning

Luijten¹,

Chennakeshava²,

Eldar³

et al. 2023

Ultrasound in Medicine & Biology

View full text Add to dashboard Cite

Section: Post-processing and Interpretationmentioning

confidence: 99%

Ultrasound Signal Processing: From Models to Deep Learning

Luijten¹,

Chennakeshava²,

Eldar³

et al. 2023

Ultrasound in Medicine & Biology

View full text Add to dashboard Cite

“…Adaptive beamforming by deep learning (ABLE) [4] has a lower computational cost than minimum variance beamforming and has been applied to improve ultrasound image quality in non-contrast enhanced images for arrays with regularly spaced elements and has been used jointly with deep learning-based localization [5] to improve the localization of microbubbles in contrast enhanced ultrasound. Inspired by this we hypothesize that ABLE can be used to improve contrast enhanced imaging with a sparse array.…”

Section: Introductionmentioning

confidence: 99%

3-D contrast enhanced ultrasound imaging of an in vivo chicken embryo with a sparse array and deep learning based adaptive beamforming

Ossenkoppele

Wei

Luijten

et al. 2022

2022 IEEE International Ultrasonics Symposium (IUS)

Self Cite

View full text Add to dashboard Cite

3-D contrast enhanced ultrasound enables better visualization of inherently 3-D vascular geometries compared to an intersecting plane. Additionally, it would allow the application of motion correction techniques for all directions. Both contrast detection and motion correction work better on high-frame rate data. However high-frame rate 3-D ultrasound imaging with dense matrix arrays is challenging to realize. Sparse arrays alleviate some of the limitations in cable count and data rate that fully populated arrays encounter, but their increased level of secondary lobes negatively impacts image contrast. Meanwhile the use of unfocused transmit beams needed to achieve highframe rates negatively impacts resolution. Here we propose to use adaptive beamforming by deep learning (ABLE) to improve the image quality of contrast enhanced ultrasound images acquired with a sparse spiral array. We train the neural network on simulated data and evaluate simulated images and in vivo images of an ex ovo chicken embryo. ABLE improved resolution compared to delay-and-sum (DAS) and spatial coherence (SC) beamforming on the simulated and in vivo data. The qualitative improvements persist after histogram matching, indicating that the image quality improvement of the ABLE images was not purely due to dynamic range stretching.

show abstract

“…Youn et al (2021) take this one step further, and combine the imagedomain LISTA architecture by Van Sloun et al with the ABLE beamforming architecture by Luijten et al (2020), training the joint network end-toend. The authors show that this outperforms non-joint optimization, and that ABLE learns to accommodate the downstream localization problem addressed by LISTA.…”

mentioning

confidence: 99%

Ultrasound Signal Processing: From Models to Deep Learning

Luijten¹,

Chennakeshava²,

Eldar³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Medical ultrasound imaging relies heavily on high-quality signal processing algorithms to provide reliable and interpretable image reconstructions. Hand-crafted reconstruction methods, often based on approximations of the underlying measurement model, are useful in practice, but notoriously fall behind in terms of image quality. More sophisticated solutions, based on statistical modelling, careful parameter tuning, or through increased model complexity, can be sensitive to different environments. Recently, deep learning based methods have gained popularity, which are optimized in a datadriven fashion. These model-agnostic methods often rely on generic model structures, and require vast training data to converge to a robust solution.A relatively new paradigm combines the power of the two: leveraging datadriven deep learning, as well as exploiting domain knowledge.

show abstract

Model-based Deep Learning on Ultrasound Channel Data for Fast Ultrasound Localization Microscopy

Cited by 4 publications

References 21 publications

Ultrasound Signal Processing: From Models to Deep Learning

Ultrasound Signal Processing: From Models to Deep Learning

3-D contrast enhanced ultrasound imaging of an in vivo chicken embryo with a sparse array and deep learning based adaptive beamforming

Ultrasound Signal Processing: From Models to Deep Learning

Contact Info

Product

Resources

About