- PhD Project Title
- Integrating flow imaging into patient-specific models following myocardial infarction
I graduated from Leeds university in 2019 with a BSc in Mathematics. During my undergraduate degree, I studied a range of applied mathematics modules, with a particular interest in fluid dynamics. For my final year project, I investigated the uses of computational modelling in applied mathematics, focussing on topics such as the use of finite difference methods for solving PDEs.
My main interests lie in the development and application of physics-informed machine learning models, specifically physics-informed neural networks (PINNs). I am also interested in super-resolution and acceleration of phase-contrast magnetic resonance imaging technologies (MRI).
My PhD project is focused super-resolution of 4D-flow MRI—a medical imaging modality that is used to reconstruct time-dependent velocity fields in the cardiovascular system—with particular focus on flow within the left ventricle (LV) of the heart. Flow imaging data are corrupted by uncertainty arising from noise and low spatio-temporal resolution, which is exacerbated in the LV due to the extreme deformation of the endocardium throughout the cardiac cycle, reducing the accuracy of clinically-relevant quantities. To improve the predictive capabilities of the 4D-flow MRI, super-resolution methods have been explored to de-noise and upsample data in space and time. Whilst a variety of approaches have been investigated in the literature, we use a PINN-based model for this application. Through weak regularisation, our model predictions are constrained to satisfy the underlying physics, allowing us to work with the spare and noisy data.
Why I chose the CDT in Fluid Dynamics
My ultimate goal is to enter a research role in industry, requiring strong research skills and an in depth knowledge of a particular subject. The CDT presents an opportunity to achieve both of these whilst developing other skills along the way. Having interests across science and mathematics, the breadth of research topics also appealed to me, and the opportunity to work in a multidisciplinary team will hopefully allow me to explore interests away from my mathematical background.
Conference and workshop presentation
- September 2022 - Oral presentation, VPH 2022 conference, Porto, Portugal
- February 2023 – Delivered SciML Workshop on PINNs, Leeds
- March 2023 - Oral presentation, LIFD: Data-driven methods in fluid mechanics workshop, Leeds, UK
- June 2023 - Poster presentation, IPMI 2023 conference, Bariloche, Argentina
- F. Shone, N. Ravikumar, T. Lassila, et al. Deep Physics-Informed Super-Resolution of Cardiac 4D-Flow MRI, proceedings from the 28th conference on Information Processing in Medical Imaging (IPMI), 2023, read the paper using its DOI.
- J. Zhang, Y. Zhao, F. Shone, et al. Physics-Informed Deep Learning for Musculoskeletal Modeling: Predicting Muscle Forces and Joint Kinematics From Surface EMG, in IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 31, pp. 484-493, 2023, read the paper using its DOI.
Linked in Profile: Find out more about me on my Linked In profile.