Artificial intelligence in molecular imaging: from machine to deep learning

During my PhD, I learned, assessed, and applied several aspects, and new potential applications of AI in molecular imaging: from ML image classification, prediction of disease outcome, prediction of response to therapy, to DL segmentation. Finally, during the period at the Universitatsspital of Zurich (USZ) I also accomplished, in collaboration with the Swiss Federal Institute of Technology (ETH), an innovative study (in submission) regarding the prediction of PET volumes from MRI images, assessing simultaneous PET/MRI. In this thesis, I will describe the main results of the abovementioned studies published during my PhD time following an anatomical and computational order.