Accelerated 4D MRI for the assessment of aortic motion

Marfan syndrome (MFS) is a genetic connective tissue disorder that increases the risk of life-threatening aortic complications such as aneurysm, dissection, and rupture. Clinical decisions are typically based on aortic diameter, yet dissections can occur before reaching surgical thresholds. This highlights the need for novel biomarkers to identify high-risk individuals earlier. Aortic motion and deformation, influenced by arterial stiffness and vessel wall biomechanics, may serve as such biomarkers. While cardiovascular magnetic resonance imaging (MRI) has improved assessment of aortic morphology, its four-dimensional (4D) potential for capturing dynamic behavior throughout the cardiac cycle remains underutilized.
This thesis focused on advancing 4D MRI methods for improved analysis of aortic motion. Specifically, it aimed to: 1) optimize the acquisition and reconstruction of a 3D time-resolved bright blood MRI sequence, and 2) develop postprocessing tools to quantify aortic dynamics, with the ultimate goal of finding novel biomarkers for aortic remodeling through (future) studies in cohorts of MFS patients and healthy volunteers.
Chapter 2 details the development of a 4D MRI pipeline using a free-running balanced steady-state free precession (bSSFP) sequence, demonstrating reproducible aortic diameter and displacement measurements in healthy volunteers. Chapter 3 presents a convolutional neural network for automated aortic segmentation, enabling assessment of dynamic motion. Chapter 4 introduces technical optimizations to improve image quality at 3 Tesla. Chapter 5 applies the pipeline in MFS patients and healthy volunteers, revealing biomechanical differences and correlations with known markers of arterial stiffness. Chapter 6 explores localized pulse wave velocity as a biomarker, though variability highlights the need for refinement.