Abdominal aortic aneurysms are pathological dilations of the aorta that constitute the most frequent structural aortic disease in humans. The prevalence is higher for males and increases with age. The main complication is a rupture of the aneurysm, associated with high mortality due to massive internal bleeding.

To prevent rupture, elective repair of aneurysms at high risk needs to be performed. This can be done via open surgery or via minimally invasive endovascular aneurysm repair (EVAR), during which stent grafts that exclude the aneurysm sac from blood flow are placed in the aorta. This procedure is aided by 2D fluoroscopy and digital subtraction angiography (DSA).

As a preoperative CT is generally available, image fusion with the live fluoroscopic images can be used to provide additional 3D information and reduce the use of contrast agent during the intervention. However, patient movement and deformations due to devices positioned in the aorta may cause inaccuracies in the overlay. This project aims to identify and address these deformations and provide an accurate and adapting overlay of 3D information during the procedure.

The current focus is on the automatic detection of relevant structures in fluoroscopic images that may influence the deformation of the aorta and the corresponding modelling of the deformation in the overlay.

This project is in cooperation with Siemens Healthcare GmbH, Forchheim.