No replication of direct neuronal activity–related (DIANA) fMRI in anesthetized mice

Abstract

Direct imaging of neuronal activity (DIANA) by functional magnetic resonance imaging (fMRI) could be a revolutionary approach for advancing systems neuroscience research. To independently replicate this observation, we performed fMRI experiments in anesthetized mice. The blood oxygenation level–dependent (BOLD) response to whisker stimulation was reliably detected in the primary barrel cortex before and after DIANA experiments; however, no DIANA–like fMRI peak was observed in individual animals’ data with the 50 to 300 trials. Extensively averaged data involving 1050 trials in six mice showed a flat baseline and no detectable neuronal activity–like fMRI peak. However, spurious, nonreplicable peaks were found when using a small number of trials, and artifactual peaks were detected when some outlier-like trials were excluded. Further, no detectable DIANA peak was observed in the BOLD-responding thalamus from the selected trials with the neuronal activity–like reference function in the barrel cortex. Thus, we were unable to replicate the previously reported results without data preselection.