Deconvolution for Hyperpolarized Magnetic Resonance Imaging

Hyperpolarized metabolic imaging is translating into clinical practice. However, the desire to obtain chemical, spatial and temporal information combined with the finite and non-renewable exogenous magnetisation provided by hyperpolarized techniques often results in imaging sequences with a poor point-spread function. Here we compare multiple deconvolution algorithms to improve apparent reconstructed resolution, as is commonly performed in PET imaging. We show that deconvolution can significantly improve the reconstructed resolution of hyperpolarized images.