Share Email Print

Proceedings Paper

Quantitative photoacoustic oximetry using convolutional neural networks (Conference Presentation)
Author(s): Kevin Hoffer-Hawlik; Austin Van Namen; Geoffrey P. Luke

Paper Abstract

Using spectroscopic photoacoustic imaging to quantitatively measure blood oxygenation saturation (sO2) is a difficult problem which requires prior tissue knowledge and costly computational methods. We have developed a convolutional neural network with a U-Net architecture to estimate the sO2 from spectroscopic photoacoustic data. The network was trained on Monte Carlo simulated spectroscopic PA data and predicted sO2 with only 4.49% error, an accuracy much higher than that of a linear spectral unmixing baseline. These results suggest that precise quantitative measurements of sO2 deep in tissue is attainable using machine learning approaches.

Paper Details

Date Published: 6 March 2020
Proc. SPIE 11240, Photons Plus Ultrasound: Imaging and Sensing 2020, 112402A (6 March 2020); doi: 10.1117/12.2545197
Show Author Affiliations
Kevin Hoffer-Hawlik, Thayer School of Engineering (United States)
Austin Van Namen, Thayer School of Engineering at Dartmouth (United States)
Geoffrey P. Luke, Thayer School of Engineering at Dartmouth (United States)

Published in SPIE Proceedings Vol. 11240:
Photons Plus Ultrasound: Imaging and Sensing 2020
Alexander A. Oraevsky; Lihong V. Wang, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?