Characterization of cerebral aneurysm morphology : development of methods and techniques in an open-source framework

The increased rate of detection of unruptured cerebral aneurysms has left clinicians with a major dilemma on whether to treat the discovered lesions by means of surgical or endovascular techniques or conservatively monitor them, the ideal situation being to treat only those aneurysms that have been diagnosed to be at high risk of rupturing. Although the precise mechanisms at the base of the pathology occurrence have not been identified yet, given the focal nature of the disease hemodynamics is widely considered as one of its main determinants. Three dimensional geometric characterization of cerebral aneurysms and their parent vasculature has also attained a major role as a powerful tool for interpreting and predicting blood flow features as well as the commonest mean to assess lesions risk of rupture in clinical settings. Both image-based hemodynamics and geometry quantification have been extensively applied to the stratification of aneurysms according to their clinical outcome; regrettably no definitive conclusions have been reached or reliable rupture risk indexes determined. The objective of this thesis is the development of computational frameworks for the image-based hemodynamic modeling and the 3D geometric characterization of patient-specific vascular districts hosting cerebral aneurysms in the context of largescale population studies. Specific features for the developed algorithms and tools are their robustness, operator-independence and automaticity. Crucial topics related to the cerebral aneurysms investigation, such as the neck section plane identification, the 3D morphologic characterization of the sac and the quantification of the spatial relationships between sac and close vasculature are specifically addressed. All the techniques for hemodynamic modeling and geometric features retrieval are part of an open source project, the Vascular Modeling Toolkit, in an attempt to encourage spreading and standardization of definitions, as well as support applied researchers in their work.