Varian Medical Systems (United Kingdom)

The speed of multidimensional NMR spectroscopy can be significantly increased by drastically shortening the customary relaxation delay between scans. The consequent loss of longitudinal magnetization can be retrieved if 'new' polarization is transferred from nearby spins. For correlation spectroscopy involving heteronuclei (X=13C or 15N), protons not directly bound to X can repeatedly transfer polarization to the directly bound protons through Hartmann-Hahn mixing. An order of magnitude increase in speed has been observed for the 600 MHz two-dimensional HMQC spectra of amikacin and strychnine using this technique, and it also reduces the noisy F1 ridges that degrade many heteronuclear correlation spectra recorded with short recovery times.

SummaryMagnetic resonance (MR) techniques have revolutionized the assessment of brain development and injury in the newborn. MR studies at 3 T or above are becoming widespread; however, given the high cost and complexity of high‐field MR systems, it is important to assess their benefits and limitations for clinical practice.MR spectroscopy (MRS) studies at high field carry potential benefits such as improved signal‐to‐noise ratio (SNR) and chemical shift dispersion compared with 1.5 T. However, safety is a very important consideration as radiofrequency (RF) tissue heating and acoustic noise also increase with field strength. Furthermore, increased chemical‐shift dependence of the position of the volume of interest (VOI), shorter transverse and longer longitudinal relaxation times constitute additional challenges.We describe the first in‐vivo, proton MRS studies of newborn human brain at 4.7 T. Nine infants with neonatal encephalopathy, all ventilated and with continuous physiological monitoring, were studied. We used L‐FOCI localization, mostly with a VOI centred on the thalami.The absence of 4.7 T results from healthy infants made it impossible to interpret pathological metabolite peak‐area ratios. However, concentration ratios and concentrations should be field‐strength independent and interpretable using control values acquired at lower field. Of eight infants with concentration‐ratio measurements, seven had low [N‐acetylaspartate (NAA)]/[choline (Cho)] and three increased [lactate]/[NAA]. Of the two infants with ‘concentration’ measurements both had low [NAA] and one reduced [creatine].We used an adult coil and consequent RF power restrictions had repercussions for the minimum magnetization recovery time (TR). A smaller, neonatal coil should improve SNR and, with less RF power, relax TR constraints thereby enabling fuller exploitation of the potential of high‐field MRS.

Background. The potential benefits of high-field magnetic resonance (MR) include increased signal, contrast, and spectral/spatial resolution, however there are specific safety and cost implications. Aims. (i) to explore the feasibility of MR imaging (MRI) and proton (1H) MR spectroscopy (MRS) at 4.7T of the newborn brain; (ii) to assess specific advantages and disadvantages of 4.7T versus low-field MR. Methods. Informed parental consent and ethical approval were obtained. Nine ventilated infants were studied at a median corrected gestational age 41+4 weeks using a whole-body 4.7T MR system with continuous monitoring (skin temperature, pulse oximetry, ECG). Acoustic noise was attenuated with incubator-lining, ear-plugs and mini-muffs. MRI included: T1- and T2-weighted single and multi-echo spin-echo, and 3D gradient-echo. Quantitative T1 and T2 maps were calculated where possible. A modified PRESS sequence was used for 1H-MRS localised to an 8mm3 thalamic voxel (echo times 144 and 288ms). Sequences were programmed to keep RF heating at less than half the recommended adult maximum levels. Results. Physiological indices were stable and no adverse events occurred. As expected, brain water T1 was higher and T2 lower at 4.7T versus lower fields. The T1-weighted images displayed poor contrast, however high spatial resolution T2-weighted imaging with a 4.7T-optimised fast spin-echo technique delivered remarkable contrast and structure differentiation (figure). 1H-MRS detected brain metabolites including lactate. Conclusions. Safety considerations at high-field limit sequence flexibility and rapidity of data acquisition. Nevertheless our initial experience demonstrates the feasibility of scanning sick newborn infants at 4.7T: The quality of T2-weighted images was significantly improved, whilst T1-weighted images and 1H-MRS were not superior at high vs low-field. Further research on the application of high-field MR to the developing brain is imperative as exploitation of specific features and use of dedicated coils may improve the diagnostic and prognostic accuracy of MR techniques.