Whole brain MP2RAGE-based mapping of the longitudinal relaxation time at 9.4T.

Mapping of the longitudinal relaxation time (T1) with high accuracy and precision is central for neuroscientific and clinical research, since it opens up the possibility to obtain accurate brain tissue segmentation and gain myelin-related information. An ideal, quantitative method should enable whole brain coverage within a limited scan time yet allow for detailed sampling with sub-millimeter voxel sizes. The use of ultra-high magnetic fields is well suited for this purpose, however the inhomogeneous transmit field potentially hampers its use. In the present work, we conducted whole brain T1 mapping based on the MP2RAGE sequence at 9.4T and explored potential pitfalls for automated tissue classification compared with 3T. Data accuracy and T2-dependent variation of the adiabatic inversion efficiency were investigated by single slice T1 mapping with inversion recovery EPI measurements, quantitative T2 mapping using multi-echo techniques and simulations of the Bloch equations. We found that the prominent spatial variation of the transmit field at 9.4T (yielding flip angles between 20% and 180% of nominal values) profoundly affected the result of image segmentation and T1 mapping. These effects could be mitigated by correcting for both flip angle and inversion efficiency deviations. Based on the corrected T1 maps, new, 'flattened', MP2RAGE contrast images were generated, that were no longer affected by variations of the transmit field. Unlike the uncorrected MP2RAGE contrast images acquired at 9.4T, these flattened images yielded image segmentations comparable to 3T, making bias-field correction prior to image segmentation and tissue classification unnecessary. In terms of the T1 estimates at high field, the proposed correction methods resulted in an improved precision, with test-retest variability below 1% and a coefficient-of-variation across 25 subjects below 3%.

Accurate phosphorus metabolite images of the human heart by 3D acquisition-weighted CSI.

Fourier imaging modalities suffer from significant signal contamination between adjacent voxels, especially when the spatial resolution is comparable to the size of the anatomical structures. This contamination can be positive or negative, depending on the spatial response function and the geometry of the object. Such a situation arises in human cardiac (31)P chemical shift imaging (CSI). Acquisition-weighted CSI reduces this contamination substantially, which is demonstrated by comparing conventional CSI to Hanning-weighted 3D (31)P-CSI experiments in 13 healthy volunteers at 2 T. The nominal spatial resolution and the total number of scans were identical for both experiments. The improved spatial response function of the acquisition-weighted experiment led to a significantly (P < 0.0001) higher myocardial PCr/ATP ratio (2.05 +/- 0.31, mean +/- SD, N = 33, corrected for saturation and blood contribution) compared to the conventional CSI experiment (1.60 +/- 0.46). This is explained by the absence of negative contamination from skeletal muscle, which also resulted in an increase of the observed SNR (from 5.4 +/- 1.4 to 7.2 +/- 1.4 for ATP). With acquisition-weighted CSI, metabolic images with a nominal resolution of 16 ml could be obtained in a measurement time of 30 min. After correction for the inhomogeneous B(1) field of the surface coil, these images show uniform ATP distribution in the entire myocardium, including the posterior wall.

Beyond k-space: spectral localization using higher order gradients.

Chemical shift imaging (CSI) often suffers from the inconvenient shape of its spatial response function (SRF), which affects both localization and signal-to-noise ratio. Replacing the magnetic field gradients for phase encoding by higher order magnetic fields allows a better adjustment of the SRF to the structures in the sample. We combined this principle with the SLOOP (spectral localization with optimal pointspread function) technique to simultaneously obtain spectra from several arbitrarily shaped compartments within a sample. Linear combinations of the fields of the shim coils are used to generate the pulsed fields for phase encoding. Their shapes are matched to the given sample geometry by numerical optimization. Using this method, spectra from a phantom were obtained that show a higher signal-to-noise ratio and a strongly reduced contamination compared to an equivalent CSI experiment.

Theoretical evaluation and comparison of fast chemical shift imaging methods.

Magnetic resonance chemical shift imaging (CSI) is becoming the method of choice for localized NMR spectroscopic examinations, allowing simultaneous detection of NMR spectra from a large number of voxels. The main limitation of these methods is their long experimental duration. A number of fast CSI experiments have been presented, promising to reduce that duration. In this contribution the criteria for evaluating and optimizing the sensitivity of fast CSI experiments are elaborated. For a typical experiment in the human brain, the performance of various methods is compared. While conventional CSI provides optimal sensitivity per unit time, it is shown in which circumstances fast sequences allow a shorter experimental duration. Using these results, the best method for any experimental requirements can be selected.