Abstract:In this paper, a new beamformer which combines the eigenspace-based minimum variance (ESBMV) beamformer with the Wiener postfilter is proposed for medical ultrasound imaging. The primary goal of this work is to further improve the medical ultrasound imaging quality on the basis of the ESBMV beamformer. In this method, we optimize the ESBMV weights with a Wiener postfilter. With the optimization of the Wiener postfilter, the output power of the new beamformer becomes closer to the actual signal power at the imaging point than the ESBMV beamformer. Different from the ordinary Wiener postfilter, the output signal and noise power needed in calculating the Wiener postfilter are estimated respectively by the orthogonal signal subspace and noise subspace constructed from the eigenstructure of the sample covariance matrix. We demonstrate the performance of the new beamformer when resolving point scatterers and cyst phantom using both simulated data and experimental data and compare it with the delay-and-sum (DAS), the minimum variance (MV) and the ESBMV beamformer. We use the full width at half maximum (FWHM) and the peak-side-lobe level (PSL) to quantify the performance of imaging resolution and the contrast ratio (CR) to quantify the performance of imaging contrast. The FWHM of the new beamformer is only 15%, 50% and 50% of those of the DAS, MV and ESBMV beamformer, while the PSL is 127.2dB, 115dB and 60dB lower. What is more, an improvement of 239.8%, 232.5% and 32.9% in CR using simulated data and an improvement of 814%, 1410.7% and 86.7% in CR using experimental data are achieved compared to the DAS, MV and ESBMV beamformer respectively. In addition, the effect of the sound speed error is investigated by artificially overestimating the speed used in calculating the propagation delay and the results show that the new beamformer provides better robustness against the sound speed errors. Therefore, the proposed beamformer offers a better performance than the DAS, MV and ESBMV beamformer, showing its potential in medical ultrasound imaging.

Deep Learning Based Minimum Variance Beamforming for Ultrasound Imaging.

Self-Adaptive Beamforming Method Based on Plane Wave Ultrasound Imaging

A Joint Method of Coherence Factor and Nonlinear Beamforming for Synthetic Aperture Imaging with a Ring Array

Coherence-based Phase Aberration Correction and Beamforming for Ring-Array Ultrasound Imaging

Adaptive Ultrasound Beamforming using Deep Learning

Adaptive and Compressive Beamforming Using Deep Learning for Medical Ultrasound

A Low-Complexity Beamformer for Ultrasound Imaging Based on Sub-Beamformer and Multi-Apodization With Cross-Correlation

A Fully Convolutional Neural Network for Beamforming Ultrasound Images

Beamforming through regularized inverse problems in ultrasound medical imaging

High dynamic range ultrasound beamforming using deep neural networks

Improving the Quality of Monostatic Synthetic-Aperture Ultrasound Imaging Through Deep-Learning-Based Beamforming

Fast Adaptive Beamforming Combined with Multiple Apodization in Ultrasound Plane Wave Imaging

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

Ultrasound deep beamforming using a multiconstrained hybrid generative adversarial network

Joint Subarray Coherence and Minimum Variance Beamformer for Multitransmission Ultrasound Imaging Modalities

Multibeam Minimum Variance Beamforming for Ring Array Ultrasound Imaging.

Universal Deep Beamformer for Variable Rate Ultrasound Imaging

Beamforming-integrated neural networks for ultrasound imaging

Experimental Validation of Ultrasound Beamforming with End-to-End Deep Learning for Single Plane Wave Imaging

Eigenspace-based Minimum Variance Beamformer Combined with Wiener Postfilter for Medical Ultrasound Imaging.

Sparse Convolutional Beamforming for Ultrasound Imaging