1st Supervisor: Pablo Lamata, King’s College London
2nd Supervisor: Andrew King, King’s College London
Industrial Co-Supervisor: Dr Alberto Gomez, Ultromics
Clinical Supervisor and Co-Supervisor: Dr Gerry Carr-White, King’s College London
Aim of the PhD Project:
- To design, develop and deploy physics informed machine learning solutions that estimate the mechanical parameters of the heart, and that unveil the stiffness parameters that will create a paradigm change in the management of conditions such as heart failure.
Lay Summary:
Heart failure is a condition where the heart is not able to meet the demands of the body. Patients with heart failure require numerous visits through the healthcare units, and thus generates a wealth of data that ends up scattered in different information systems. This research project aims to deliver a better management of heart failure by focusing on the patient journey, where the system is able to ensure the right patients get the right treatment or technology at the right time, in the right place. The specific objective is to unveil one of the hidden characteristics of the beating heart: the stiffness of its walls. By enabling the descriptive analysis of how the stiffness of the heart changes through health and disease, this project will deliver a step change in the understanding and characterization, and thus management, of heart failure patients.
The PhD student will be supervised leading technologists and cardiologists from King’s College London. Prof. Lamata and Dr. King are leading the construction of the digital twin of the heart of each patient. In health care the digital twin denotes the vision of a comprehensive, virtual tool that integrates coherently and dynamically the clinical data acquired over time for an individual using mechanistic and statistical models. King’s Health Partners (KHP) is leading the Value Based Healthcare Programme, where Dr. Carr-White is the lead on the cardiovascular care pathway with a specific focus on heart failure.
The PhD project work will specifically tackle the challenge of how to personalise the digital twin to each patient (its mechanical material parameters) given the limited available routine clinical data. The project will develop AI technologies that learn from the previous examples (neural networks) while constrained by the fundamental laws of physics (physics informed). The project will work with the widely prevalent but challenging ultrasound data, where a direct observation of deformation is available, and where an indirect estimation of pressure can be inferred. The project will be supported (in-kind, and an industrial placement) by an Oxford spin-out, Ultromics, that is expert in the development of AI tools for the analysis of ultrasound data.
References
- [Arthus21] https://www.sciencedirect.com/science/article/pii/S002199912100259X
- [Buoso21] https://www.sciencedirect.com/science/article/pii/S1361841521001122
- [Burnhope19] https://pubmed.ncbi.nlm.nih.gov/30744444/
- [Corral20] https://academic.oup.com/eurheartj/crossref-citedby/5775673
- [Lamata16] https://link.springer.com/article/10.1007/s10439-015-1439-8
- [Webb18a] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6128173/
- [Webb18b] https://www.sciencedirect.com/science/article/abs/pii/S0167527317345011