Aortic dissection is a life threatening traumatic disease in which the media of the aorta "delaminates" (i.e., it splits in two) with blood entering the created cavity. Depending on the location of the dissection, lesions are categorized as type A (involving the ascending aorta) or type B (involving the descending aorta). The extent of the dissection is highly variable; in some patients, the lesion is confined to a relatively narrow zone, with coagulation of the blood and a stabilization and healing of the wound. In other patients, however, the dissection may span the entire aorta, with a peeled off aorta starting near the aortic arch and extending beyond the iliac bifurcation. Very often, there is blood flow within the formed cavity (called the false channel) with blood entering/leaving the false channel via one or more tears (allowing direct communication between the true aortic lumen and the false channel). After some time, the false channel may (partly) coagulate, which seems to depend on whether or not side branches of the aorta are supplied with blood via the false or the real lumen. Treatment of patients with aortic dissection can be based on medication alone (typically to lower the blood pressure) and/or may involve surgery and placment of a stent graft, where it is often the aim to seal off the entry to the false lumen to stimulate coagulation of blood (and subsequent healing). In a substantial amount of patients, however, treatment with a stent graft does not stop the progression of the disease, and the aorta continues to dilate posing a severe risk of rupture and death of the patient. The mechanisms determining whether or not a dissection will stabilize and heal after treatment, or continue to progress are not very well understood. The extent of coagulation of the false channel seems to play a role, but it is also believed that the mechanical interaction of a stent graft with the dissected aorta and its environment may play a role. The figures below show some CT scans of patients with aortic dissection, illustrating the variability of the disease.
Thoracic aorta with false lumen (blue arrow) and coagulated blood (green arrow).
Fully dissected thoracic aorta
Before and after placement of a stent graft
Through our collaborations with UZ Gent and the University Hospital in Düsseldorf, we obtained a very large dataset of 96 patients with aortic dissection, who had a CT scan upon arrival in the hospital and follow-up scans in the following weeks/months/years. Patients were treated in different ways, with some patients receiving medication only, while others were treated with a stent graft. In some patients, the lesion healed over time or is under control and stable, while in other patients the dissection continues to dilate after receiving a stent graft. As it is, up to now, unclear which patients will benefit from this stent graft placement, we want to develop computational tools that enable to predict the acute and long-term patient-specific outcome of this treatment in patients with a type B dissection. One of the tools that are essential to achieve this goal, is a reliable model of the stent graft placement itself, allowing us to predict the deployed stent graft configuration. As it is envisaged, on the long term, to combine this model with other computational tools (e.g. model of the blood flow, thrombus formation, …), computational cost is important as well.
Therefore, it is the aim of this thesis to examine which degree of complexity is required to reliably model the deployed stent graft configuration in a patient with a type B aortic dissection.
In order to do so, 3D finite element modelling techniques (Abaqus) will be used to set up models of the stent graft deployment and the aortic dissection with varying degrees of complexity regarding the aortic wall geometry and material properties, the stent graft material properties and deployment location and methodology. Each of the models will be set up based on patient-specific CT scans and will be complemented with literature/clinical data. Although the obtained stresses and strains are relevant as well, this study will mainly focus on the deployed stent graft geometry and the computational cost of the model. The predicted deployed stent graft geometry of the different models will be compared amongst each other and to the real deployed geometry, that can be obtained from the CT-scans by doing a segmentation in Mimics (Materialise).