Weakly supervised brain tumour segmentation using generative adversarial networks

This project is an EPSRC-funded PhD studentship focused on the development of weakly/self-supervised deep learning approaches for segmentation of lesions in medical images. Using widely available cross-sectional measurements (e.g. RANO/RECIST), this project aims to produce adversarial learning approches for robust segmentation in the absence of large amounts of annotation labelmaps.