1st Supervisor: Alistair Young, King’s College London
2nd Supervisor: Jack Lee, King’s College London
Clinical Champion: Bijan Modarai, King’s College London
Industrial Supervisor: Raquel Sanchez, Clinical Programs Manager, Cook Medical
Aim of the PhD Project:
- Train a machine learning algorithm to characterise the aortic arch and proximal descending aorta from CT images
- Determine the proximal landing zone in patients receiving TEVAR
- Develop in silico simulations of total endovascular solutions for preserving the supra-aortic trunks with a focus on the left subclavian artery
Project Description / Background:
Endovascular aortic repair is a relatively common procedure performed in over 450 patients per year at King’s Health Partners hospitals. Typically, a stent is introduced via a catheter placed in the femoral artery and deployed in the thoracic aorta to repair a dissection or aneurysm or stenosis. Successful repair requires an optimal proximal landing zone (PLZ). Often this requires covering an important artery such as the left subclavian artery (LSA). Studies suggest that the LSA needs to be covered in up to 80% of cases to allow an acceptable PLZ for thoracic endovascular aortic repair (TEVAR) in patients with type B aortic dissection (Van der Weijde et al. J Endo Therapy 2017). LSA blood flow can impact stroke and paraplegia after TEVAR but the absence of a good off-the-shelf endovascular solution for incorporating the LSA, particularly during acute cases, means that preservation rates remain low.
This project will develop tools for optimizing the PLZ in patients who are at risk of compromised LSA. Firstly, machine learning medical image analysis methods will be trained on several hundred cases obtained from hospital databases. This will be used to gain an objective understanding of anatomical nuances related to the PLZ (in healthy and diseased aorta). These tools will then be applied to a large, multi-centre cohort using CT images of patients who have undergone TEVAR involving the aortic arch/proximal descending aorta. Finally, biomechanical simulations will be developed to examine the positioning of stent grafts in the distal arch and proximal descending thoracic aorta after deployment, including the how the anatomy of the PLZ influences stent deployment.
This project would suit an engineer interested in machine learning and image analysis for optimal design of medical devices.
Figure 1: CT angiogram of aorta prior to thoracic endovascular aneurysm repair (TEVAR) and after TEVAR. The stent graft has achieved a seal in the proximal and distal healthy aorta and the aneurysm is sealed as evidenced by lack of contrast within it.