Training Deep Network Ultrasound Beamformers With Unlabeled In Vivo Data
- 30 December 2021
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Imaging
- Vol. 41 (1) , 158-171
Abstract
Conventional delay-and-sum (DAS) beamforming is highly efficient but also suffers from various sources of image degradation. Several adaptive beamformers have been proposed to address this problem, including more recently proposed deep learning methods. With deep learning, adaptive beamforming is typically framed as a regression problem, where clean ground-truth physical information is used for training. Because it is difficult to know ground truth information in vivo , training data are usually simulated. However, deep networks trained on simulations can produce suboptimal in vivo image quality because of a domain shift between simulated and in vivo data. In this work, we propose a novel domain adaptation (DA) scheme to correct for domain shift by incorporating unlabeled in vivo data during training. Unlike classification tasks for which both input domains map to the same target domain, a challenge in our regression-based beamforming scenario is that domain shift exists in both the input and target data. To solve this problem, we leverage cycle-consistent generative adversarial networks to map between simulated and in vivo data in both the input and ground truth target domains. Additionally, to account for separate as well as shared features between simulations and in vivo data, we use augmented feature mapping to train domain-specific beamformers. Using various types of training data, we explore the limitations and underlying functionality of the proposed DA approach. Additionally, we compare our proposed approach to several other adaptive beamformers. Using the DA DNN beamformer, consistent in vivo image quality improvements are achieved compared to established techniques.Keywords
Funding Information
- NIH (S10OD016216-01, R01EB020040)
- NSF (IIS-175099)
- Data Science Institute at Vanderbilt University
This publication has 41 references indexed in Scilit:
- Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification2015
- A model and regularization scheme for ultrasonic beamforming clutter reductionIEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, 2015
- Reverberation Clutter from Subcutaneous Tissue Layers: Simulation and in Vivo DemonstrationsUltrasound in Medicine & Biology, 2014
- Ultrasonic multipath and beamforming clutter reduction: a chirp model approachIEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, 2014
- Short-lag spatial coherence of backscattered echoes: imaging characteristicsIEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, 2011
- Eigenspace-based minimum variance beamforming applied to medical ultrasound imagingIEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, 2010
- Benefits of minimum-variance beamforming in medical ultrasound imagingIEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, 2009
- Broadband minimum variance beamforming for ultrasound imagingIEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, 2009
- The van Cittert–Zernike theorem in pulse echo measurementsThe Journal of the Acoustical Society of America, 1991
- Multilayer feedforward networks are universal approximatorsNeural Networks, 1989