Self-attention feature pyramid network for multimodal medical image segmentation

My research project aims to utilise the data of the CBTN dataset to demonstrate the proof of concept of a novel model for automated medical anomaly detection in magnetic resonance images and to incorporate it into a bigger dataset designed to investigate the generalisability of the proposed model for multi-institutional data. This project constitutes a subproject of my research I am conducting with my supervisors from the University of Technology Sydney (Prof. Paul Kennedy and A/Prof. Daniel Catchpoole) and physicians at the Children’s Hospital at Westmead (Dr. Robert Goetti, Dr. Dinisha Govender and Prof. Stewart Kellie). By providing additional auxiliary information to physicians in routine clinical assessment, decision-making can be improved, which ultimately leads to an improvement in patient care.