Cerebrovascular reactivity (CVR) is conventionally assessed by comparing discrete cerebral blood flow measurements obtained prior to and following a hemodynamic stimulus. Blood-oxygen-level-dependent (BOLD)-MRI provides potentially continuous, dynamic CVR characterization when coupled to stimuli such as acetazolamide (ACZ), but poor signal-to-noise generally limits its analysis to an analogous comparison of arbitrarily defined terminal BOLD signals relative to baseline. We present a novel framework incorporating temporal and spatial denoising to pre-condition the dynamic time-signal course for robust voxel-level maximal and terminal CVR, and novel CVR time-to-peak and related kinetic features, disambiguating confounders related to potentially non-optimal terminal CVR assignment in conventional use.
This abstract and the presentation materials are available to members only; a login is required.