Task-Evoked Functional Activation and Coupling With CSF Flow Detected in the Human Brain With Ultrashort Echo Time fMRI at 7 T.

Ultrashort echo time (UTE) and zero echo time (ZTE) techniques have been shown to generate robust functional MRI (fMRI) contrast of hemodynamic origin from the inflow of unperturbed spins into the imaging volume under local radiofrequency transmission. As such, they are ideally suited not only to map the dynamics of cerebral blood flow (CBF) during intense neuronal activity but also the dynamics of cerebrospinal fluid (CSF). The goal of this work was to pilot the use of UTE-fMRI for human studies of brain activation and neurofluid dynamics. We thus conducted fMRI studies at 7 T with a transmit/receive head coil using a slab-selective UTE sequence on 13 human participants during a visual task. Spatio-temporally matched gradient-echo echo planar imaging (GE-EPI) data were also acquired for qualitative comparison purposes, and respiratory and cardiac signals were measured to quantify multiple physiological metrics. Functional brain activations were analyzed using a general linear model at single subject and group levels. Moreover, time-courses from the whole cortex, carotid arteries, and fourth ventricle were correlated between each other, and physiological contributions to these ROI signals were evaluated with a linear mixed model and a relative importance computation. Robust and reproducible functional activations were detected with UTE-fMRI in the visual cortex across participants. Although the functional contrast-to-noise ratio was higher with GE-EPI than with UTE, the temporal signal-to-noise ratio was lower, and group-level statistical power and activation patterns of UTE maps were similar to those obtained with GE-EPI. In addition, the UTE task-evoked signal in CSF was negatively correlated with those in the whole cortex and in the carotid arteries and was primarily driven by the stimulus paradigm. We conclude that UTE-fMRI can be used not only for functional studies of the human brain but also for assessing the relationship between hemodynamic and CSF signals, which can help elucidate brain homeostatic processes.