| Cerebrovascular disease is a major cause of stroke. Digital subtraction angiography (DSA), the gold standard for definitive diagnosis and evaluation of cerebrovascular disease, is invasive and involves clinically significant risk of complications. Magnetic resonance imaging (MRI) as a non-invasive and safe imaging modality plays an increasingly important role in the diagnosis, treatment planning and follow-up evaluation of cerebrovascular disease. In addition to anatomy, hemodynamic information is also essential to understanding the pathophysiology, development and progression of cerebrovascular disease.;Ultrasound-based sonography and two-dimensional (2D) phase-contrast MRI (PC-MRI) are two major means to measure cerebral vascular flow and velocities. However, sonography is limited by low reproducibility and inadequate insonation window when measuring vessels distal to the circle of Willis. 2D PC-MRI only includes planar hemodynamic information and requires repetitive 2D plane placement to measure multiple vessel locations. Four-dimensional (4D) flow MRI (also termed "time-resolved 3D phase-contrast MRI" or "flow-sensitive 3D velocity mapping") provides reliable flow measurement consistent with 2D PC-MRI and offers additional benefits including full volumetric coverage of the vasculature of interest, 3D visualization of complex flow patterns and retrospective flow quantification at any desired vessel locations within the imaging volume.;The purpose of this work was to develop and optimize 4D flow MRI protocols for intracranial applications and to develop data analysis methods for comprehensive intracranial hemodynamic assessment in normal subjects and patients with cerebrovascular disease.;First, 4D flow MRI protocols with optimized pulse sequence parameters for intracranial studies were developed. A three-step data analysis workflow was developed for processing the 4D flow MRI data. The first step is preprocessing, including noise masking and correction of phase errors from eddy currents, Maxwell terms and velocity aliasing as well as calculation of a 3D phase-contrast MR angiogram (PC-MRA). The second step is to visualize cerebral 3D blood flow patterns using time-resolved and time-integrated 3D pathlines. The third step is to quantify blood flow and velocities in manually positioned 2D analysis planes at specified vessel locations.;Next, the established 4D flow MRI protocols and data analysis tools were applied to systematically characterize intracranial hemodynamics in normal subjects and patients with cerebrovascular disease. First, 4D flow MRI was combined with cardiac 2D PC-MRI to characterize age-related changes of normal cerebral and cardiac flow in a cohort of volunteers with ages spanning infancy to middle adulthood. After that, 4D flow MRI was applied to evaluate intracranial hemodynamic alterations in three cohorts of patients with cerebral arteriovenous malformation, vein of Galen aneurysmal malformation and intracranial atherosclerotic disease, respectively. |
