Accelerator Architecture for Plane-Wave Ultrasound Image Reconstruction in Fourier Domain

Pooriya Navaeilavasani;Daler Rakhmatov
{"title":"Accelerator Architecture for Plane-Wave Ultrasound Image Reconstruction in Fourier Domain","authors":"Pooriya Navaeilavasani;Daler Rakhmatov","doi":"10.1109/OJUFFC.2025.3530395","DOIUrl":null,"url":null,"abstract":"Ultrafast ultrasound imaging based on coherent plane-wave compounding (CPWC) enables very high data acquisition rates in the order of thousands of frames per second. This capability allows the user to capture and characterize fast-changing dynamics of blood flow or tissue motion, thus facilitating advanced biomedical diagnostics. Fast data acquisition should be supported by high image reconstruction rates, which translates into significant computational demands. To address this issue, several state-of-the-art hardware accelerators for CPWC image reconstruction, or beamforming, have been reported in the literature. They primarily target time-domain methods based on delay-and-sum (DAS) beamforming. For the first time, this article proposes a novel hardware architecture for accelerating Fourier-domain image reconstruction, based on an efficient migration technique from geophysics. Our FPGA implementation of one specific architectural instance achieves the reconstruction throughput of 1,380 frames per second (without compounding), where each complex-valued “analytic” image frame consists of <inline-formula> <tex-math>$2048\\times 128~64$ </tex-math></inline-formula>-bit data samples. The presented work also aims to motivate further research into hardware support for Fourier-domain migration. This technique is asymptotically faster than conventional DAS beamforming; however, its efficient hardware realization is challenging, partly due to its relatively large memory footprint.","PeriodicalId":73301,"journal":{"name":"IEEE open journal of ultrasonics, ferroelectrics, and frequency control","volume":"4 ","pages":"231-246"},"PeriodicalIF":0.0000,"publicationDate":"2025-01-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10843301","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"IEEE open journal of ultrasonics, ferroelectrics, and frequency control","FirstCategoryId":"1085","ListUrlMain":"https://ieeexplore.ieee.org/document/10843301/","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

Abstract

Ultrafast ultrasound imaging based on coherent plane-wave compounding (CPWC) enables very high data acquisition rates in the order of thousands of frames per second. This capability allows the user to capture and characterize fast-changing dynamics of blood flow or tissue motion, thus facilitating advanced biomedical diagnostics. Fast data acquisition should be supported by high image reconstruction rates, which translates into significant computational demands. To address this issue, several state-of-the-art hardware accelerators for CPWC image reconstruction, or beamforming, have been reported in the literature. They primarily target time-domain methods based on delay-and-sum (DAS) beamforming. For the first time, this article proposes a novel hardware architecture for accelerating Fourier-domain image reconstruction, based on an efficient migration technique from geophysics. Our FPGA implementation of one specific architectural instance achieves the reconstruction throughput of 1,380 frames per second (without compounding), where each complex-valued “analytic” image frame consists of $2048\times 128~64$ -bit data samples. The presented work also aims to motivate further research into hardware support for Fourier-domain migration. This technique is asymptotically faster than conventional DAS beamforming; however, its efficient hardware realization is challenging, partly due to its relatively large memory footprint.
傅立叶域平面波超声图像重建的加速器结构
基于相干平面波复合(CPWC)的超快超声成像可以实现每秒数千帧的高数据采集率。该功能允许用户捕获和表征血流或组织运动的快速变化动态,从而促进先进的生物医学诊断。快速的数据采集需要高图像重建率的支持,这意味着大量的计算需求。为了解决这个问题,文献中已经报道了几种用于CPWC图像重建或波束形成的最先进的硬件加速器。他们主要针对基于延迟和(DAS)波束形成的时域方法。本文首次提出了一种新的硬件架构,用于加速傅里叶域图像重建,该架构基于地球物理的有效迁移技术。我们的FPGA实现了一个特定的架构实例,实现了每秒1,380帧的重建吞吐量(无复合),其中每个复值“分析”图像帧由$2048 × 128~64$位数据样本组成。提出的工作也旨在激励进一步研究硬件支持的傅里叶域迁移。该技术比传统的DAS波束形成渐近快;然而,其高效的硬件实现是具有挑战性的,部分原因是其相对较大的内存占用。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信