ABSTRACT
BACKGROUND AND OBJECTIVES: We compared functional MRI acquisition methods of sparse temporal acquisition (STA) and continuous acquisition (CA) to estimate the effect of MRI scanner background noise (SBN) on blood oxygen level-dependent (BOLD) activation of cortical and subcortical auditory centers during auditory stimulation. SUBJECTS AND METHOD: Fourteen healthy subjects (eight males, age 30.6+/-4.7 years) were presented with classical music in a block paradigm (36 s on/off) in two STA [repitition time (TR)=12 s, 60 volumes] and two CA (TR=2 s, 360 volumes) functional MRI sessions. To account for the sample size difference, an additional volume-matched continuous dataset (CAm) was generated by matching CA to 60 volumes of STA. A group-level analysis based on BOLD activation maps was performed. Percent signal change (PSC), T-statistic values and signal variability in cortical and subcortical auditory regions of interest (ROIs) were calculated from individual activation maps and compared between the STA, CA, and CAm. RESULTS: The group analysis showed activation in the primary and secondary auditory cortices in all datasets. However, the activation of subcortical auditory centers above the accepted threshold was only observed in STA. STA (less SBN) showed higher PSCs and T-statistic values in all ROIs except planum temporale when compared to CAm. However, there was no difference in signal variability among the datasets. CONCLUSION: Our results suggest that SBN should be considered as a significant confounder in auditory-evoked functional MRI studies particularly in the activation of subcortical auditory centers, and that STA can be an effective imaging method for reducing the effect of SBN.