Summary Brief periods of neural activity trigger a mixture of vascular and metabolic responses with somewhat stereotypical dynamics. This hemodynamic response function (HRF) is heavily used in functional magnetic resonance imaging (fMRI). While the HRF is usually assumed to be spatially invariant, detailed measurements show that its dynamics exhibit significant temporal variability; such variability confounds standard linear analysis of fMRI measurements. Nonetheless, the variability likely reflects changing interactions between blood flow and oxygen metabolism that could provide critically useful information about brain function. Thus, it is of strong interest to better understand the physics and physiology of the HRF. We developed and validated the Arterial Impulse Model (AIM) that combines a novel flow description with compartmental, 1D, convection-diffusion oxygen transport. We propose further development and validation of the model, coupled with detailed investigation of the HRF. Experimentally, we use multi-sensory stimulus protocols to evoke the HRF broadly across the brain, and evaluate the responses using high-resolution arterial spin labeling and fMRI to examine the character and variability of these responses across the brain. We will expand our model to explain the biomechanical basis of this HRF variability in a far richer framework than previous models, with the potential to transform fMRI from a rough neural correlate to a quantitative measurement of cerebral blood flow and metabolism. We will also develop a clinical application of our whole- brain HRF methods to evaluate early Alzheimer’s Disease (AD) or Mild Cognitive Impairment. Our simple, hour-long fMRI evaluation has strong potential to provide a prospective biomarker of AD pathology.