Candidate:
William
Pitman
Title:
Receiver
Channel
Recovery
in
High-Frame-Rate
Ultrasound
Imaging
using
Branched
Convolutional
Neural
Networks
Date:
July
26,
2022
Time:
2:00
pm
Place:
online
Supervisor(s):
Yu,
Alfred
Abstract:
High frame rate ultrasound (HiFRUS) can be used to track dynamic events in the body, and its integration into compact ultrasound scanners could enable use of the paradigm in more remote and austere healthcare settings. While desired, the high data rates and dedicated receiver electronics required for HiFRUS acquisitions make their implementation on compact systems difficult. Reduction of the number of receiving channels in a HiFRUS system can alleviate constraints related to data rate and receiver electronics, however, this reduction will degrade beamformed image quality. To circumvent this issue, a novel receiver channel recovery framework was developed. A branching encoder-decoder convolutional neural network (CNN) architecture was designed to facilitate channel recovery after multiple rates of downsampling, and 3 CNN instances were trained to recover omitted radiofrequency (RF) channels from angled plane wave acquisitions after downsampling rates of 2-times, 3-times, and 4-times. To evaluate the utility of the trained CNNs, recovered RF data was used for ultrasound image formation using delay-and-sum beamforming, and for coherent compounding of beamformed images. When the trained recovery frameworks were applied to downsampled acquisitions of an in vitro point target and an in vivo carotid, inclusion of CNN-inferred RF data removed spatial aliasing artifacts from the beamformed images, recovering their underlying structure. The proposed framework may be used to improve image quality on more compact and inexpensive HiFRUS systems that operate with a subset of receiving channels.