MultiPRESS: Deep learning for super-resolution 4D Flow MRI
||MultiPRESS: Deep learning for super-resolution 4D Flow MRI|
||David Marlevi <firstname.lastname@example.org>|
||2023-10-01 – 2024-04-01|
Medical imaging has come to revolutionize the way we diagnose, monitor, and treat disease. This is particularly true in the field of cardiovascular care, where a range of imaging modalities are used to interrogate anatomy and function using non-invasive means. However, a range of clinical challenges remain where available image techniques are hindered by inherent method limitations (restricted resolution; excessive noise; etc.). The incorporation of machine learning techniques and data-driven algorithms into clinical image research is however starting to fundamentally change what can be analysed, and resolved, opening for individualized patient-care and directly improved clinical risk stratification.
As part of a recent ERC-StG, the PI and team are now using dedicated machine learning techniques to push the limits of hemoydnamic full-field imaging. Specifically, we are exploring super-resolution neural networks to enhance the spatiotemporal coverage of 4D Flow magnetic resonance imaging (MRI); a novel imaging techniques providing 3D blood flow coverage across the cardiovasculature and over time. With this work already part of an ongoing snic application, this proposal is submitted as an replacement of the prevoius allocation, targeting ensemble (bagging, boosting, stacking) and generative models to enhance generalizability. In particular, the added proposal is necessary, considering the computational resources required to scale up to combination of base/meta networks.