Project Summary Cerebral arteriovenous malformation (AVM) is a congenital cerebrovascular malformation that predisposes to vascular aneurysm, hemorrhagic stroke, and seizures. This typically occurs between ages 10 – 40, making AVM an unpredictable, potentially devastating condition in young, previously asymptomatic patients. The current clinical standard in AVM evaluation and surgical planning is to derive morphology and subjective flow parameters from digital subtraction angiography (DSA) – an effective but invasive technique that uses radiation and contrast. The goal of this project is to develop an optimized 4D flow MRI sequence for imaging AVM hemodynamics, in order to provide quantitative metrics of net flow, flow distribution, peak velocity, and pressure drop across the AVM, and to track these parameters throughout AVM treatment. 4D flow MRI is a method developed in the Markl lab to obtain temporally- and spatially-resolved three- dimensional flow velocity data from a contrast-free MRI scan. In adapting this approach for neurovascular imaging and AVM specifically, the aim is to reduce total imaging and processing time while increasing spatial resolution to account for small vessels, increasing dynamic range of measured velocities to accommodate AVM physiology, and maintaining clinical utility of flow images. This project will not only include advanced MRI sequence development, but also rely on close collaboration with clinicians to ensure that the sequence is optimized to provide types of information that can be clinically useful in AVM evaluation and treatment. The first aim is development of a sequence with an appropriate undersampling technique and data collection parameters, as well as a network-based, streamlined data processing method. Design criteria are chosen to enable both visualization of main feeding vessels of an AVM and evaluation of pulsatile flow within a cardiac cycle. The second specific aim is validation of the developed sequence and data processing method in vitro, with a model system with flow properties similar to actual AVMs, and in vivo, in healthy volunteers. Using an existing MRI-compatible flow pump with physiological flow profile and a custom 3D-printed flow chamber (MRI phantom), we will test the performance of the developed sequence and compare the resulting measurements to ground truth values calculated using the phantom geometry. Then, 10 healthy volunteers will be recruited for a test-retest reproducibility study to assess inter-study mean differences and limits of agreement. The third aim is validation of the developed sequence and data processing method in AVM patients, via a small pilot study. 10 adult AVM patients will be recruited at Northwestern Memorial Hospital. Patients will be scanned prior to first treatment as well as after each embolization step. Through feedback from collaborating physicians, the potential clinical utility of the new information obtained from the scan will be evaluated and opti...