CEST peaks are easy to detect at ultra-high-field strengths due to high signal and spectral separation. However, spectral coalescence and line broadening makes modeling of CEST effects at clinical field strengths (<=3T) a challenge. In this proof-of-concept study of super-resolution CEST imaging, the underlying spectral features of 3T Z-spectra were predicted using a neural network trained on 9.4T data. Applying the neural network to untrained volunteer and patient data acquired at 3T resulted in the expected contrast in healthy gray and white matter and tumor tissue in Z-spectra and APT, NOE, and MT CEST maps.
This abstract and the presentation materials are available to members only; a login is required.