Although there have been countless medical advances in the past years, there are still many diseases that are difficult to diagnose or require invasive procedures to diagnose. This project aims to rigorously evaluate the applicability of machine learning in improving the diagnostic accuracy of various diseases, including cardiomyopathies and metabolic liver diseases. We will use large sets of MRI data matched with corresponding patient data to develop markers of disease that can be recognised on imaging. We expect that designing and stringent testing and revising of this algorithm will take approximately three years. Our hope is that by having a way to effectively and efficiently diagnose disease without invasive procedures that need to be performed by specialists in large healthcare centres will substantially improve access and quality of healthcare.
