University of Houston

Endovascular treatment planning of intracranial aneurysms requires accurate quantification of their geometric parameters. Currently, the geometry of intracranial aneurysms is typically quantified manually based on three-dimensional (3D) Digital Subtraction Angiography (DSA) images. Since the reproducibility of manual measurements is not guaranteed and their accuracy depends on the experience of the treating physician, there is a need for computer-aided estimation of these parameters.

In this thesis, we present a framework for computer-aided quantification of the aneurysmal geometry that is useful for endovascular treatment of intracranial aneurysms. In particular, we first perform a vessel segmentation and extract the surface of the segmented vessel. We then separate the aneurysm from its parent vessels and localize its neck. Finally, we estimate clinically relevant geometric parameters of the aneurysm, including the neck length, dome height and maximum diameter. We examine several methods for vessel and aneurysm segmentation, present their advantages and limitations, and propose a semi-automatic method. The results indicate that there is good agreement between the geometric parameters estimated by the proposed approach and manual measurements obtained by two experienced interventional neuroradiologists.