This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. Introduction: Fetal functional MRI (f-fMRI) may be useful for differentiating between normal and compromised fetal well-being in pregnancies at risk for intrauterine growth restriction (IUGR). Despite the relatively small size of the fetal head, conventional pulse sequences require long readouts for a field of view (FOV) that encompasses the maternal abdomen in order to avoid aliasing [1]. We compare the performances of conventional and reduced FOV (rFOV) pulse sequences for fMRI of the fetal brain. Methods: The rFOV gradient echo (GRE) pulse sequence commences with two frequency-selective fat-saturation RF pulses (260Hz offset, bandwidth = 220Hz) and crusher gradient sets. Next, the spins are prepared by a two-dimensional spatially-selective pulse [2] and an adiabatic half-sech fast passage pulse to achieve outer volume saturation (OVS) before slice selective excitation. For the adiabatic pulse, the effective B1 initially points along the transverse net magnetization vector of the previously selected spins. As B1 amplitude decreases, and off-resonance increases, the effective B1 sweeps up toward its final longitudinal orientation. A standard slice selective 90x generates signal. The readout follows a spiral-out trajectory in k-space. Subjects were late term pregnant women at risk for IUGR. The fMRI stimuli consisted of 15s blocks of classical music alternating with 15s blocks of silence; and maternal breathing 1min. blocks of oxygen separated by varying intervals of medical air. 128 and 210 time frames were acquired with TE/TR=70ms/2s, resulting in a 4:15min. and 7min. protocols, respectively, on a 1.5T scanner (GE) receiving with the manufacturer s 8-channel cardiac array coil. fMRI series for each stimulus were acquired with a conventional spiral GRE acquisition and with OVS for a rFOV (24cm) readout. The respective conventional/rFOV imaging parameters were: FOV=44/24cm, number of pixels = 128/40. The slice thickness and spacing were 5mm and 1mm for both sequences. The shorter readout of the rFOV sequence (11.9ms) allowed the gradient slew rate to be lowered from 150T/m/s to 30T/m/s, resulting in a 21.4ms readout duration with scanner noise reduction (90dB from 97dB) and SNR gain of 15%. Data post-processing consisted of selective receiver coil exclusion, time frame alignment by rigid body transformation (SPM2), interpolation of time frame slices the global intensity of which varied by more than 3 standard deviations from the series mean, and visual inspection of time series in order to delete misaligned frames. Activation is estimated as the cross-correlation coefficients (cc) between the time series and stimulus covariate functions. Acknowledgements: This work was supported in part by the NCI Training Grant T32 CA09695 and NIH R33 CA88205NIH, and the Center of Advanced MR Technology at Stanford (P41RR09784), and IR01EB0 02771. References: 1. Fulford, J., S. H. Vadeyar, et al. (2004). 'Fetal brain activity and hemodynamic response to a vibroacoustic stimulus.' Human brain mapping. 22(2): 116-21. 2. Pauly J, Nishimura D, Macovski A. A k-space analysis of small-tip-angle excitation. J Magn Reson 1989;81:43-56.
Showing the most recent 10 out of 446 publications
