Artificial Intelligence Enhanced Ultrasound Flow Imaging at the Edge

Ultrasound flow imaging has long been used for cardiovascular diagnostics. Color Doppler imaging (CDI) is the predominant ultrasound flow imaging mode, but its diagnostic value is hampered by aliasing artifacts that limit the range of detectable blood velocities. Here, we present the first demonstra...

Full description

Saved in:
Bibliographic Details
Published in:IEEE MICRO 2022, Vol.42 (6), p.1-11
Main Authors: Nahas, Hassan, Huver, Sean, Yiu, Billy Y. S., Kallweit, Chris M., Chee, Adrian J. Y., Yu, Alfred C. H.
Format: Article
Language:English
Subjects:
Acceleration
Aliasing
Artificial intelligence
Doppler effect
Edge computing
Frames (data processing)
Frames per second
Graphics processing units
Image edge detection
Imaging
Pipelines
Pipelining (computers)
Real time
Real-time systems
Ultrasonic imaging
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Ultrasound flow imaging has long been used for cardiovascular diagnostics. Color Doppler imaging (CDI) is the predominant ultrasound flow imaging mode, but its diagnostic value is hampered by aliasing artifacts that limit the range of detectable blood velocities. Here, we present the first demonstration of how edge artificial intelligence (AI) can enable real-time CDI with aliasing resistance. Specifically, graphical processing unit (GPU) acceleration and AI-ready edge computing hardware have been leveraged to realize the first end-to-end CDI processing pipeline that involves AI-based aliasing correction. Performance results show that, using our edge AI engine, aliasing-resistant CDI frames with threefold velocity detection range can be generated at a real-time frame rate of 25 fps (for raw datasets with 192 channels, 12-bit data resolution, and 25 MHz sampling rate). Overall, edge AI can critically improve the real-time visualization quality of ultrasound flow imaging and, in turn, potentially transform its bedside application value.
ISSN:0272-1732
1937-4143
DOI:10.1109/MM.2022.3195516