Direct mapping of neural activity via glutamate-weighted magnetic resonance imaging

Conventional functional magnetic resonance imaging (fMRI) indirectly maps neural activity via blood oxygenation level dependent (BOLD) contrast.1 Directly probing glutamate, the primary excitatory neurotransmitter, during task-related stimulation offers a more precise neuroimaging approach. Chemical exchange saturation transfer (CEST), a molecular MRI technique, can provide insights into brain metabolites,2,3 including glutamate.4 While prior studies have attempted to apply CEST in fMRI contexts—either to detect glucose consumption in rats5 or to probe pH changes in the human brain6 (with the latter failing to detect significant changes)—none of these efforts successfully addressed the confounding influence of BOLD signal fluctuations on the acquired CEST contrast during task activation. Furthermore, to date, no study has reported task-evoked activation maps based on dynamic glutamate changes detected by CEST-MRI.

Here, we are the first to propose and implement a dynamic signal model that integrates both BOLD and CEST effects. Derived from the analysis of temporal response amplitude, we introduce a metric potentially detecting changes in glutamate concentration during neural activity, along with corresponding glutamate-weighted activation maps. Simulation results based on this model show strong agreement with experimental results during block-design visual tasks in the human brain on a 3 Tesla scanner, but only when incorporating increased glutamate concentration during the stimulation state. Furthermore, the glutamate-weighted activation maps obtained in our experiments demonstrate a more precise localization of visual cortex regions compared to the classical BOLD contrast activation maps. Our findings provide a potential explanation for the mechanism underlying CEST-fMRI and underscore the potential of this imaging modality to directly map neural activity by leveraging glutamate concentration detection.