MRI has great clinical values to distinguish soft-tissues without contrast or radiation. By using the hybrid-modality information from MRI and PET, here we developed deep learning method to synthesize metabolic activity mapping from contrast-free multi-contrast MRI images. Trained on clinical datasets, we demonstrated the feasibility to estimate metabolic biomarker from contrast-free MRI and validated on both FDG-PET/MRI and Amyloid-PET/MRI in-vivo datasets. This technique can be used for more efficient, low-cost, multi-tracer functional imaging, exploring anatomy-function relationship, visualizing new bio-markers and improving the workflow for both MRI and PET/MRI.
This abstract and the presentation materials are available to members only; a login is required.