Machine learning methodology for training cardiac digital twins

Project description

Digital twins are mathematical models of physical phenomena trained on data. They can be used to aid understanding and guide decision making, and are increasingly being used across industry and in medical applications. Cardiac digital twins are models of patient hearts that can be used to guide patient-specific treatment. This project will develop machine learning methodology to aid the training of cardiac digital twins. It forms part of a larger project, joint with Imperial College London, looking at moving digital twin technology towards where it can be trialled in a formal clinical trial.

The focus will be on developing neural network tools to quickly personalise digital twins to new patients. Mathematically, the aim is to solve Bayesian inference problems for complex computer models at speed.