Computational Framework to predict re-operative needs in cases of Tetralogy of Fallot and Coarctation of the Aorta

Congenital Heart Disease (CHD) is a heart condition resulting from an abnormality in heart structure or function that is present at birth. In the UK alone, about 4,600 babies are born with this condition each year. Computational Fluid Dynamics (CFD), or in silico modelling, has had a profound impact on cardiovascular medicine in the past decade and will be used in this PhD project to assess its potential as an assistive tool in diagnosis and treatment of two CHD conditions, namely Tetralogy of Fallot (ToF) and Coarctation of the Aorta (CoA). This interdisciplinary project will aim to use engineering simulation tools to develop a practical workflow to help identify haemodynamic features that indicate need for reoperation in the case of ToF and as a tool to predict hypertension in CoA. The proposed PhD project will involve 4 phases:

Phase1: Structural and flow data will be extracted from CT, echo and serial MR scans in a series of child patients with either repaired ToF or repaired CoA. Age range targetting period of rapid deterioration of RV function or increase in blood pressure.

Phase2: Following an initial validation of the methodology, CFD analysis to be used to determine haemodynamic influence on post-surgical morphological recovery. Features that might correlate with need for re-operation, which the models will quantify, include excessive haemodynamic load on tissue (leading to fibrosis), excessive loads on valves (leading to degeneration and/or structural deterioration), and thrombosis on valves (risk of which can be assessed using standard flow derived metrics).

Phase3: A detailed morphological characterisation of the geometry from extracted 4D image sets will be performed to identify post-repair evolution in volume, sites of wall-remodelling, fibrosis, valve-annulus dimension, degree of regurgitation. Changes to be parameterised and compared with CFD results from the Phase2 to identify direct causation.

Phase4: Additional data from the registry such as genetic characterisation and modifiable/non-modifiable risk factors will be incorporated into a data-mining framework based on Bayesian Gaussian modelling. CFD models in TOF and CoA will be built to generate pilot data regarding the potential of the models to predict
haemodynamic deterioration using the longitudinal data collected during this project.

Computational Details: The project will focus on the development of open-source software towards the preparation of a tool for use in clinical environments. It is envisaged that both Finite Volume and Lattice Boltzmann methods will be explored along with a range of methods to facilitate the incorporation of moving boundaries and flow-structure interaction.