Spectroscopic photoacoustic (sPA) imaging can be used to map blood oxygen saturation (sO2) within tissue. Its accuracy, however, is degraded deep in tissue by wavelength-dependent optical attenuation. We have developed a convolutional neural network to simultaneously estimate the sO2 and segment blood vessels from sPA data. The network was trained on Monte Carlo simulated sPA data and predicted sO2 with 9.31% median pixel error. The network was then retrained on experimental photoacoustic images of cow blood with median prediction error of 4.38%. These results suggest that precise quantitative measurements of sO2 deep in tissue are attainable using machine learning approaches.
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.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.