Perfusion MRI for Multi-site Studies of Brain Function

点击翻译按钮获取中文摘要

Arterial Spin Labeling (ASL) perfusion MRI shows promise as a widely available and quantitative measure of resting brain function that can be used as a biomarker for the neural correlates of psychiatric and neurologic diseases, for the measurement of drug effects in the brain, and ultimately for diagnosis and treatment monitoring in individual patients. Although ASL perfusion MRI technologies are now implemented on most MRI scanner platforms, uncertainty in the community on the performance differences of various technical implementations and how to best use the technology across centers and MRI scanner platforms has limited its dissemination into routine use in clinical research. Building upon our successful development of quantification methods and standard sequences across platforms in the prior funding cycle, we propose to focus on the key issues limiting multi-center studies with ASL. Our first aim is to measure the relative sensitivity of different implementations of ASL by comparing their reproducibility and their response to 2 test interventions, citalopram or alprazolam administration. The result of this aim will be a quantitative calculation of power for detection of regional effects and how the choice of implementation will affect the power and required sample size. The second aim is to develop quality assessment methods using customized image acquisitions and the construction of a perfusion phantom. Quality assurance is a key element of imaging studies across sites, but there are no established methods for testing ASL perfusion performance. In our final aim, we target the characterization and reduction of variable perfusion signal induced by changes in brain activity unrelated to study interventions, so-called physiological noise. We will determine whether performing a moderately demanding vigilance task during the ASL scan will help control the subject's mental state and reduce variability without excessively stimulating particular regions of the brain. We will also study resting fluctuations in perfusion induced by network activity in the brain to determine if identifying and removing these fluctuations during image processing improves reproducibility. This aim will also determine if the amplitude of resting fluctuations is reflective of resting perfusion. Since resting brain fluctuations as measured by blood oxygenation sensitive MRI are increasingly being used as an indicator of resting function, establishing a relationship between fluctuations and average resting activity will address an outstanding question in functional imaging. Achievement of these aims will accelerate and improve the use of ASL as a biomarker for brain function in disease and will greatly improve the design of numerous planned and active multi-site studies employing ASL.