Cone-Beam Computed Tomography For Neuro-Angiographic Imaging
MetadataShow full item record
Accurate and timely diagnosis and treatment of head trauma incidents, particularly ischemic stroke, rely on confident visualization of brain vasculature and accurate measurements of flow characteristics. Due to their low cost, high spatial resolution, small footprint, and adaptable geometry, cone-beam computed tomography (CBCT) systems have shown to be valuable for point of care applications that require timely diagnosis. However, image quality in CBCT scanners are potentially limited by slow rotation speed and are therefore challenged in the context of hemodynamic neurovascular imaging, as in CT Angiography (CTA) and CT Perfusion (CTP). This thesis details the advancement of neuro-angiographic imaging in CBCT systems using prior-image based reconstruction methods such as Reconstruction-of-Difference (RoD) to provide accurate visualization of cerebral vasculature despite slow image acquisition and/or sparse projection data. The performance of a novel CBCT system for high quality neuro-angiographic imaging was assessed using RoD, a prior-image-based reconstruction method. A digital simulation framework is presented that involves dynamic vasculature models and an accelerated forward projection method. Results indicate feasibility for CBCT angiography attained with a 5 second scan time, despite a limited number of projections (“sparse” projection data). Preliminary results in CTP imaging demonstrate the potential for perfusion parameter estimation and warrant future research using a 3D printed vessel phantom developed in this work.