Complex Convolutional Neural Networks for Ultrafast Ultrasound Imaging Reconstruction From In-Phase/Quadrature Signal

• Published in: IEEE transactions on ultrasonics, ferroelectrics, and frequency control ferroelectrics, and frequency control, 2022-02, Vol.69 (2), p.592-603
• Main Authors: Lu, Jingfeng, Millioz, Fabien, Garcia, Damien, Salles, Sebastien, Ye, Dong, Friboulet, Denis
• Format: Article
• Language: English
• Subjects: Artificial neural networks; Beamforming; Cardiology and cardiovascular system; Complex convolutional neural networks; Computer architecture; Computer Science; Convolutional codes; Convolutional neural networks; deep learning; diverging wave; Fluid Dynamics; Human health and pathology; Image Processing, Computer-Assisted - methods; Image quality; Image reconstruction; Imaging; in-phase/quadrature signal; Life Sciences; Medical Imaging; Neural networks; Neural Networks, Computer; Physics; Quadratures; Radio frequency; Radio signals; Signal and Image Processing; Signal processing; ultrafast ultrasound imaging; Ultrasonic imaging; Ultrasonography - methods
• ISSN: 0885-3010; 1525-8955
• DOI: 10.1109/TUFFC.2021.3127916

Ultrafast ultrasound imaging remains an active area of interest in the ultrasound community due to its ultrahigh frame rates. Recently, a wide variety of studies based on deep learning have sought to improve ultrafast ultrasound imaging. Most of these approaches have been performed on radio frequency (RF) signals. However, in- phase/quadrature (I/Q) digital beamformers are now widely used as low-cost strategies. In this work, we used complex convolutional neural networks for reconstruction of ultrasound images from I/Q signals. We recently described a convolutional neural network architecture called ID-Net, which exploited an inception layer designed for reconstruction of RF diverging-wave ultrasound images. In the present study, we derive the complex equivalent of this network, i.e., complex-valued inception for diverging-wave network (CID-Net) that operates on I/Q data. We provide experimental evidence that CID-Net provides the same image quality as that obtained from RF-trained convolutional neural networks, i.e., using only three I/Q images, CID-Net produces high-quality images that can compete with those obtained by coherently compounding 31 RF images. Moreover, we show that CID-Net outperforms the straightforward architecture that consists of processing real and imaginary parts of the I/Q signal separately, which thereby indicates the importance of consistently processing the I/Q signals using a network that exploits the complex nature of such signals.