Abstract Body: Introduction Cerebral blood flow (CBF) is a major determinant of when and where within the brain stroke occurs. The brain’s demand for oxygen and nutrients makes it sensitive to pulsatility propagating from the heart. The contribution of cardiac flow to CBF is poorly understood. We present a pilot study using 4D flow MRI to sequentially measure both aortic blood flow and CBF in the same individual and our initial findings comparing blood pressure (BP), cardiac flow, and CBF in healthy volunteers.
Methods We acquired both 4D flow brain MRI and 4D flow cardiac MRI in a single session using a Siemens 3T Magnetom Vida scanner in 5 volunteers (4 male, 1 female) 31.2 ± 10.4 years old. Using a compressed sensing protocol (R=7.7) each 4D flow scan was acquired in under 15 minutes. Brain and cardiac MRI respectively used a 1.4 mm³ and 2.1 mm³ voxel size; velocity was encoded in 3D at 120 cm/s and 200 cm/s, with 15-25 cardiac phases. 4D flow brain coverage included the ICA, BA, MCA, ACA, and PCA (Fig 1). 4D flow cardiac coverage included the ascending and descending thoracic aorta (Fig 2). 4D flow data were reconstructed using Tempus Pixel (Tempus, IL) and regions of interest were defined to measure volumetric flow rate (Q). Blood flow pulsatility over the cardiac cycle was quantified using pulsatility index (PI). Pearson’s correlation coefficient (r) was calculated for statistical analysis.
Results Aorta Q correlated strongly with cerebral Q, in the ICA and BA (r=0.90), in the MCA, ACA, and PCA (r=0.70). Aorta Q correlated with PI in both the ICA and BA (r=0.81), but not within the MCA, ACA, and PCA (r=0.12). Aorta PI was moderately anticorrelated with PI throughout the cerebral arteries (r=-0.52). Brachial systolic and diastolic BP both correlated with cerebral PI (r=0.68, 0.99, respectively), while moderately anticorrelated with aorta PI (-0.66, -0.52). Pulse pressure correlated strongly with cerebral Q (r=0.85), and less with aorta Q (r=0.59). Heart rate was strongly anti-correlated with aorta PI (r=-0.82) and moderately correlated with cerebral PI (r=0.61).
Conclusion These initial findings demonstrate that flow characteristics Q and PI are similar between the aorta and major arteries feeding the brain (ICA, BA), but the effects of BP and heart rate appear to invert at more distal vessels (MCA, ACA, PCA). We plan to expand this technique to investigate the connection between cardiac and cerebral flow in patients with neurovascular disease.