Vessel tractography for magnetic particle imaging angiography

Extraction of vascular structures, such as cerebral arteries from angiography data, is an important step in the detection and analysis of vessel anomalies and pathologies such as aneurysms. It is also important to create accurate and smooth surface models for the hemodynamic or flow dynamic analysis, functional assessment, interventional and surgical planning, and simulation. Extracted surface models can also be used for quantitative measurements and geometrical analysis of the tube-like structures in other domains. Manual segmentation is an exhaustive task for experts. Hence, various vasculature segmentation algorithms have been developed in the literature, and several reviews have been published [1, 2]. In this work, we propose a vessel tractography approach that would lend itself naturally with the ideal bright blood contrast of Magnetic Particle Imaging (MPI) [3-5]. We demonstrate this new vessel segmentation method on Magnetic Resonance Angiography (MRA) data, and extract a whole vessel tree using a high order tensor (HOT) vessel tractography approach inspired by Higher Angular Diffusion Imaging (HARDI) techniques [6,7].