Dual-echo EPI for non-equilibrium fMRI - implications of different echo combinations and masking procedures

Dual-echo EPI is based on the acquisition of two images with different echo times per excitation, thus allowing for the calculation of purely T2(*) weighted data. The technique can be used for the measurement of functional activation whenever the prerequisite of constant equilibrium magnetization cannot be fulfilled due to variable inter-volume delays. The latter is the case when image acquisition is triggered by physiological parameters (e.g. cardiac gating) or by the subject's response. Despite its frequent application, there is currently no standardized way of combining the information obtained from the two acquired echoes. The goal of this study was to quantify the implication of different echo combination methods (quotients of echoes and quantification of T(2)(*)) and calculation modalities, either pre-smoothing data before combination or subjecting unsmoothed combined data to masking (no masking, volume-wise masking, joint masking), on the theoretically predicted signal-to-noise ratio (SNR) of the BOLD response and on activation results of two fMRI experiments using finger tapping and visual stimulation in one group (n=5) and different motor paradigms to activate motor areas in the cortex and the brainstem in another group (n=21). A significant impact of echo combination and masking procedure was found for both SNR and activation results. The recommended choice is a direct calculation of T(2)(*) values, either using joint masking on unsmoothed data, or pre-smoothing images prior to T(2)(*) calculation. This method was most beneficial in areas close to the surface of the brain or adjacent to the ventricles and may be especially relevant to brainstem fMRI.