Ultrasound imaging has been used to visualise structures that are too small to be resolved by the human eye, making it an important tool to visualise the structure of internal organs with fine detail. The cardiovascular engineering group at the University of Bath have been applying this to the heart to gauge a better understanding of how blood flows in the heart, and how it may be affected by various deformities. The heart is quite deep in the chest cavity, so the ultrasound images are susceptible to noise and interference. The aim of this project is to combine the Navier-Stokes equations that govern fluid dynamics with neural networks to reduce the effects of this noise and better resolve the physical properties of the image.
Medical imaging, Physics-informed neural networks, Machine learning for medical applications, Quantum computing, Quantum machine learning
Graduated from Imperial College London with a First Class MSci Physics with Theoretical Physics. Undergraduate Masters dissertation investigated the suitability of quantum machine learning for medical image classification.
Dr Katharine Fraser
Dr Deblina Bhattacharjee
