Deep neural networks for medical image segmentation [PhD]

Artificial intelligence keeps transforming critical diagnostic practices of healthcare sectors and has gained immense attention from researchers and practitioners. This is particularly observed in areas where larger and larger datasets are gathered under clinical protocols, or are made available by medical institutions worldwide. The later case is the main reason why liver tumor diagnosis has gained such great visibility in recent years. My focus of research is on under explored areas of research that may aid in the production of more advanced methods both in the step of dataset pre-processing, with the development of pertinent generative models, to the modelling approach performance, with the proposal of novel architecture designs, whose decisions are well justified using model interpretability, and whose areas of uncertainty are information additionally extracted.