BACKGROUND AND PURPOSE: Previous studies have primarily used single-voxel t
echniques to obtain MR spectra from the neonatal brain. In this study, we a
pplied 3D MR spectroscopic imaging techniques to detect the spatial distrib
ution of MR spectroscopic imaging-detectable compounds in premature and ter
m infants. The goals were to test the feasibility of obtaining 3D MR spectr
oscopic images of newborns, assess the spatial variations of metabolite lev
els, and determine age-dependent differences in MR spectroscopic imaging da
ta.
METHODS: MR spectroscopic imaging data were acquired from nine premature (p
ostconceptional age, 30-34 weeks) and eight term (postconceptional age, 38-
42 weeks) neonates, all with normal clinical and neurologic outcomes. A spe
cialized point-resolved spectroscopy sequence with very selective saturatio
n pulses was used to select a region encompassing the majority of the brain
. Phase encoding in three dimensions was performed in a 17-minute acquisiti
on time to obtain 3D spectral arrays with a 1.0 cm(3) nominal spatial resol
ution.
RESULTS: This study showed the feasibility of detecting the 3D distribution
s of choline, creatine, and N-acetylaspartate resonances in the neonatal br
ain. Significant spectral differences were detected among anatomic location
s and between the premature and term groups.
CONCLUSION. This initial study indicates that 3D MR spectroscopic imaging o
f the neonatal brain can detect anatomic and age-dependent variations in me
tabolite levels. This technique seems to be a powerful tool to assess the m
etabolic differences between anatomic regions and to follow the changes in
cellular metabolites with brain maturation. This study also indicates the n
eed for determining topologic and age-matched normative values before metab
olic abnormalities in neonates can be accurately assessed by MR spectroscop
y.