• Corpus ID: 237605053

Deep Learning for Ultrasound Beamforming

  title={Deep Learning for Ultrasound Beamforming},
  author={Ruud van Sloun and Jong-Chul Ye and Yonina C. Eldar},
Diagnostic imaging plays a critical role in healthcare, serving as a fundamental asset for timely diagnosis, disease staging and management as well as for treatment choice, planning, guidance, and follow-up. Among the diagnostic imaging options, ultrasound imaging is uniquely positioned, being a highly cost-effective modality that offers the clinician an unmatched and invaluable level of interaction, enabled by its real-time nature. Ultrasound probes are becoming increasingly compact and… 

Figures and Tables from this paper

Ultrasound Signal Processing: From Models to Deep Learning

An overview of model-based deep learning techniques for medical ultrasound applications, which require less trainable parameters and training data than conventional neural networks, and concludes with a future perspective on these model-∗ based solutions.



Adaptive and Compressive Beamforming Using Deep Learning for Medical Ultrasound

A deep neural network is designed to directly process full or subsampled radio frequency data acquired at various subsampling rates and detector configurations so that it can generate high-quality US images using a single beamformer.

Efficient B-Mode Ultrasound Image Reconstruction From Sub-Sampled RF Data Using Deep Learning

This paper proposes a novel deep learning approach that directly interpolates the missing RF data by utilizing redundancy in the Rx–Xmit plane and results confirm that the proposed method can effectively reduce the data rate without sacrificing the image quality.

Compressed Fourier-Domain Convolutional Beamforming for Wireless Ultrasound imaging

The proposed compressed Fourier domain convolutional beamforming method can generate B-mode images using 142 times less data than DAS and can pave the way towards wireless US and demonstrate that high resolution US images can be produced using sub-Nyquist sampling and a small number of receiving channels.

Sparse Convolutional Beamforming for Ultrasound Imaging

The results demonstrate that COBA outperforms DAS in terms of resolution and contrast and that the suggested beamformers offer a sizable element reduction while generating images with an equivalent or improved quality in comparison with DAS.

Deep-Learning Based Adaptive Ultrasound Imaging From Sub-Nyquist Channel Data

This work proposes an approach for deep-learning-based reconstruction of B-mode images from temporally and spatially sub-sampled channel data, and yields high-quality B- mode images, with up to two times higher resolution than previously proposed reconstruction approaches from compressed data as well as delay-and-sum beamforming of the fully-sampling data.

Reconstruction for Diverging-Wave Imaging Using Deep Convolutional Neural Networks

A convolutional neural network (CNN) architecture, Inception for DW Network (IDNet), for high-quality reconstruction of DW ultrasound images using a small number of transmissions, and outperforming more conventional CNN architectures in terms of complexity, inference time, and image quality.

Deep Learning to Obtain Simultaneous Image and Segmentation Outputs From a Single Input of Raw Ultrasound Channel Data

Overall, the DNNs successfully translated feature representations learned from simulated data to phantom and in vivo data, which is promising for this novel approach to simultaneous ultrasound image formation and segmentation.

Deep Unfolded Recovery of Sub-Nyquist Sampled Ultrasound Images

A reconstruction method from sub-Nyquist samples in the time and spatial domain is proposed, which is based on unfolding the iterative shrinkage thresholding algorithm (ISTA), resulting in an efficient and interpretable deep network.

Learning Sub-Sampling and Signal Recovery With Applications in Ultrasound Imaging

A deep learning solution termed Deep Probabilistic Sub-sampling (DPS) is proposed, that enables joint optimization of a task-adaptive sub-sam sampling pattern and a subsequent neural task model in an end-to-end fashion.

A compressed beamforming framework for ultrafast ultrasound imaging

A CB framework, applicable to either radio-frequency or in-phase quadrature data and for both plane wave and diverging wave compounding, is described and solves a convex problem and assumes sparsity in a wavelet-based model to achieve high-quality image reconstruction from measurements acquired with only few transducer elements.