Resolving intrinsic ambiguities of the fMRI signal in the human brain

Abstract

Functional magnetic resonance imaging (fMRI) using the blood oxygenation level dependent (BOLD) signal is a standard tool in human neuroscience studies. However, the BOLD signal is physiologically complex, depending on the baseline state of the brain and the balance of changes in blood flow and oxygen metabolism in response to a change in neural activity. Interpretation of the magnitude of the BOLD response is thus problematic; specifically, group differences of the BOLD response to a standard task or stimulus, or a change after administration of a drug, could be due to a change in the neural response, neurovascular coupling, or the baseline state. While changes in oxygen metabolism can serve as a biomarker of neural activity change, reflecting the energy cost of activity, oxygen metabolism cannot be estimated from BOLD measurements alone. Here we used the effects of caffeine as a test case to show that a suite of additional noninvasive measurements, requiring no manipulation of inhaled gases, makes it possible to untangle the underlying effects of caffeine on blood flow and oxygen metabolism. After caffeine administration, the BOLD response to a standard motor task was reduced, but oxygen metabolism was not reduced; rather, caffeine reduced the balance of blood flow relative to oxygen metabolism, both in the baseline state and in response to the task. The MRI methods used required only an additional 12 minutes of scanning, can be implemented on any MRI system, and painted a more complete picture of the physiological effects of caffeine. In addition to the assessment of drug effects in the human brain, these methods make it possible to resolve ambiguities of the BOLD signal in studies of development, aging, and disease. (Duplicated from first paragraph of manuscript)