A Novel Segmentation and Navigation Tool for Endovascular Stenting of Aortic Aneurysms M. Feuerstein a , K. Filippatos a , O. Kutter a , E.U. Schirmbeck b , R. Bauernschmitt b , N. Navab a a Chair for Computer Aided Medical Procedures (CAMP), Technische Universität München, Germany b Department of Cardiothoracic Surgery, Deutsches Herzzentrum München, Germany Abstract. Endovascular stenting is a minimally invasive technique to exclude an aortic aneurysm or dissection from the circulatory system. Currently, there is no technical aid to guide the surgical staff during the intervention, except the default visualization interface provided with the CT scanner and mobile C-arm. The purpose of our proposed system is two-fold: (1) In the planning phase, a modified graph cuts algorithm automatically segments the aorta and aneurysm, so the surgical staff can choose an appropriate type of stent to match the segmented location, length, and diameter of the aneurysm and aorta. (2) During implantation of the stent graft, after a landmark based registration of CT and angiography data, the current position of the stent can be visualized in the 3D CT data set at any time. This will enhance the accuracy of the actions of the surgeon, along with a minimum use of angiography, leading to less radiation exposure and less contrast agent injection. Keywords: Minimally Invasive Surgery, Endovascular Stenting, Computer Aided Surgery, Segmentation, Navigation 1. Purpose The implantation of a stent graft, also referred to as endovascular stenting, is a minimally invasive surgical procedure that involves the exclusion of an aneurysm from the circulatory system. This is achieved by the placement of a prosthesis (the stent) that is attached to the aorta above and below the dilatation. The implant is mounted on a catheter, which is introduced from the femoral or iliac artery, until the abdominal or thoracic part of the aorta is reached, depending on the location of the aneurysm [1,2]. In the current clinical workflow, pre-operatively a high-resolution contrasted CT of the patient’s circulatory system is acquired. Using the visualization interface provided with CT scanner, the procedure is planned. The exact positions of the aneurysm and the branching vessels of the aorta need to be detected manually. Also the stent type is selected manually, based on measured length of the aneurysm and diameter of aorta. We propose a method to allow the user to select one seed point inside the aorta curvature in order to segment whole aorta and aneurysm inside a high-resolution volume. Intra-operatively, a mobile fluoroscope is used to support the surgeon in positioning the stent. Under angiography, contrast agent is applied periodically to visualize the blood vessels and the current location of stent. Since resulting X-ray image sequences are not registered to the pre-operative CT data, the high-resolution CT data is currently not used intra-operatively. A navigation purely based on two-dimensional image series leads to inaccuracies in the viewing direction, accompanied by manual re-adjustments of the angle of C-arm and acquisition of additional image series. For the successful implantation of a single stent graft, 15 to 20 X-ray image series are necessary. For every image series under angiography contrast agent is injected in the circulatory system. This frequent use of C-arm increases radiation exposure for patient as well as surgical staff. Additionally, contrast agent injection can cause adverse effects. We propose a navigation component for visualizing current stent position within pre- operative CT volume at any time. This increases the accuracy of intervention while reducing radiation exposure and contrast agent injection.