MAGNETIC RESONANCE IMAGING OF CEREBROVASCULAR REACTIVITY

Cerebrovascular reactivity (CVR) is an important indicator of cerebrovascular reserve and provides important information about vascular health in a range of brain conditions and diseases. Unlike steady-state vascular parameters, such as cerebral blood flow (CBF) and cerebral blood volume (CBV), CVR measures the ability of cerebral vessels to dilate or constrict in response to challenges or maneuvers. However, the current CVR measurement have several limitations. First, large variations in CVR across subjects and across different sessions of the same subject are often observed. Second, the current standard CVR measurement involves the use of hypercapnia challenge or acetazolamide injection which is cumbersome for routine clinical uses. Third, information in CVR MRI data has not been fully utilized. In particular, the MRI signal during a CVR experiment contains information of both neural and vascular functions. Thus CVR MRI data may be used to investigate brain functional connectivity and vascular reactivity concomitantly. This dissertation aimed to systematically investigate technical improvements of CVR MRI. First, we identified the physiological components underlying the observed variability in CVR data in both within-subject and between-subject settings. Second, we developed the approach to obtain brain resting-state functional connectivity (FC) from CVR data. Third, we developed a deep-learning model to estimate cerebrovascular reactivity (CVR) and bolus arrival time (BAT) mappings from resting-state functional MRI