Dissertation Defense of Stefanie Demirci

Title: New Approaches to Computer Assistance for Endovascular Aortic Repairs

Abstract:

With the establishment of minimally-invasive procedures for abdominal surgery, image guidance has become more and more important. The intraoperative use of medical imaging data has on the one hand increased the safety and robustness of such procedures. On the other hand, it has increased the radiation exposure. The introduction of computer based navigation systems has improved the situation to a great extent. However, they are not optimized for every application in abdominal catheter interventions.

This work presents new approaches to computer assistance for endovascular abdominal aortic repairs. In the current clinical workflow there is no technical guidance during the intervention except for two dimensional X-Ray images. An integration of preoperative three dimensional image data is further complicated by medical instruments occluding relevant anatomical structures in the images. Here, we introduce interventional registration methods that are able to handle such occlusions and further image dissimilarities. Thereby, contrast removal and disocclusion techniques are integrated within an image registration procedure. Furthermore, we show for the first time that a detection of the stent graft in the interventional image and a three dimensional recovery of its shape can be obtained at the same time. This allows image-based tracking of the prosthesis and simultaneous visualization within a volume rendering of the patient scan. All introduced methods are completely image based and do not require additional equipment to be introduced into the intervention room. The thesis is completed by presenting detailed evaluation of the methods using both synthetic images and real clinical data.

Committee Head: Prof. Dr. Johann Schlichter, TUM
1st Referee: Prof. Dr. Nassir Navab, TUM
2nd Referee: Prof. Dr. Franjo Pernus, University of Ljubljana

Date: Tuesday, April 12, 2011

Time: 9:00 (20min public presentation + approx. 30-40min non-public examination)

Location: MI 03.13.010

Technische Universität München
Fakultät für Informatik
Boltzmannstr. 3
85748 Garching b. München
Germany