Mri Assessment Of Maternal Uteroplacental Circulation In Pregnancy

Eileen Hwuang, University of Pennsylvania


Hypertensive pregnancy disorders (HPD) such as preeclampsia are highly associated with maternal vascular malperfusion of the placenta, an organ that exchanges nutrients and oxygen between the maternal circulation and the growing fetus. Adverse pregnancy outcomes are difficult to predict because there is insufficient understanding of how poor maternal arterial remodeling leads to disease. There is also a lack of reliable tools to evaluate these changes in early gestation.

The hypothesis of this dissertation was that magnetic resonance imaging (MRI) could noninvasively evaluate uteroplacental function in vivo through a combination of arterial spin labeling (ASL), 4D flow, and time-of-flight (TOF) techniques which were already effective in the evalution of other cardiovascular diseases. These flow and perfusion imaging studies were conducted on human pregnant volunteers in their second and third trimesters at 1.5T. Many of them were also examined by conventional Doppler ultrasound (US) and followed through delivery.

Flow-sensitive Alternating Inversion Recovery (FAIR) ASL MRI with background suppression was found to be feasible in detecting placental perfusion signal despite the presence of motion artifacts. An important consideration when studying placental ASL was the slow movement of maternal arterial blood in a large cavity called the intervillous space. This was a unique feature of placental anatomy which distinguished it from other organs containing capillaries. It became apparent that traditional models to estimate perfusion from MRI were no longer applicable. In this work, a statistical approach was first developed to filter out motion artifacts, followed by a coordinate transformation to better represent the lobular distribution of blood flow in the intervillous space of the placenta. The uterine arteries (UtAs) are the main maternal blood supply of the placenta and have also long been suspected to be involved in HPD, though US-based measurements have not yet been found to be highly predictive for widespread clinical use. In this work, 4D flow MRI enabled visualization of the tortuous UtAs while measuring volumetric flow rate. Its performance in predicting incidence of preeclampsia and small-for-gestational age births was comparable to Doppler US. When considering the innovative potential of 4D flow MRI to capture complex flow dynamics, this validation demonstrated the value of continuing technical development for improving HPD risk assessment. Furthermore, centerline extraction of the maternal pelvic arteries in TOF MRI, from the descending aorta to the UtAs and external iliac arteries, provided quantitative metrics to characterize the geometry including path length and curvature. Pulse wave velocity (PWV) was estimated using path length by TOF MRI and velocimetry by 2D phase contrast and 4D flow MRI with results showing sensitivity to differences between UtAs and external iliac arteries. These approaches provided physiological metrics to explore and characterize the remodeling process of the uteroplacental arteries. This dissertation demonstrates the feasibility of measuring structure and hemodynamics of the maternal vascular blood supply using non-contrast MRI that can lead to the more reliable biomarkers of adverse pregnancy outcomes needed to diagnose and treat HPD.