Metabolites such as total choline (tCho=choline+phosphocholine+glycerophosphocholine), total creatine (tCr=creatine+phosphocreatine), and total n-acetyl-aspartate (tNAA=NAA+NAAglutamate) are all intracellular, but reside within different axonal and glial cells¹. Using high b-value diffusion-weighted spectroscopy (DWS), PCho, tCr, and tNAA have been shown to exhibit non-Gaussian diffusion (non-monoexponential signal decay) in the healthy human brain, although these metabolites are distributed differently among cellular microstructures². For accurate characterization of metabolite diffusion in the intracellular space, DWS measurements require b-factors close to 20,000 s/mm² and therefore suffer from relatively low SNR. Previous studies have shown that the increased transmit field efficiency using high permittivity materials enabled significant increases in spectral SNR in the medial temporal lobe at 7T, despite a reduced receive sensitivity produced by the dielectric pads³. In this study, we investigate the effects of designing dielectric pads to enhance the B₁^- receive sensitivity enhancements in parietal white matter, in particular for low SNR applications such as DWS at extremely high b-factors.

Seven healthy volunteers (25±4 years, 4 female, 3 male) were scanned on a 7T Philips Achieva MRI scanner. The pad (15x15x1 cm³, suspended barium titanate) was placed between the volunteer's head (positioned for the parietal volume) and the 32-channel receive channel coil as shown in Fig. 1⁴. Local calibration of the transmit gain was performed by acquiring a whole-head B₁⁺ map at a spatial resolution of 2.5x2.5x5 mm³ using the dual refocused acquisition mode (DREAM) sequence⁵,TR/TE = 4.5/1.8 ms. The B₁⁺ field was calibrated within the spectroscopic VOI to produce 90° and 180° flip angles in the cases of both with and without the dielectric pad. The B₁^- receive sensitivity was first estimated using a 3D gradient-echo sequence with a tip angle of 1°, aligned with the DREAM geometry. Then, the ratio of the 3D gradient-echo image and the B₁⁺ map yielded the receive sensitivity map in arbitrary units^6,7.

Fig. 2a shows the 8 cm³ volume of interest (VOI) planned in the a) mostly parietal white matter (WM) for diffusion-weighted spectroscopy. The DWS data were acquired with a 13-interval STEAM sequence using bipolar diffusion gradients and cardiac synchronization⁸. Three orthogonal directions [1,1,-0.5], [1,-0.5,1], and [-0.5,1,1] were chosen to maximize the gradient strength for the isotropic diffusion weighting. DWS parameters were TR/TE=3000/105 ms, Δ=100 ms, δ=30 ms, τ=13 ms, and one gradient amplitude producing a b-factor of 17,794 s/mm². The isotropic DWS data were eddy-current and phase corrected using custom Matlab codes as previously described^9,10. The SNR was calculated by finding the individual peak intensities of PCho, tCr, and tNAA and comparing against the standard deviation of the residual spectra (i.e., frequency>6 ppm). The TE value was fixed at 105 ms for DWS measurements both without and with pad.

Example B₁^- receive sensitivity maps without (Fig. 2b) and with pad (Fig. 2c), show an average improvement of ~34% within the locally power-optimized volume. Fig. 3 shows diffusion-weighted spectra (b=17,794 s/mm²), demonstrating an SNR increase of ~33-48% for the metabolites. Fig. 4 shows paired comparisons of DWS SNR improvement (p=0.0002) when using the high-permittivity pad. Fig. 5 shows the pad provided a clear increase in B₁^- sensitivity (p<0.0001) as well as a mild but significant decrease in B₁⁺ variance within the localized volume (p=0.0148).

Using a single dielectric pad, the SNR for DWS data acquired from the parietal lobe increased by more than 30%, which allows for a decrease in data acquisition time of ~70% for a given SNR⁴. Our results indicate the SNR boost is mostly due to increased receive sensitivity (closer coupling of the receive array to the head). The pad also provided slightly improved local B₁ homogeneity (lower flip angle variability) within the DWS VOI, which also results in a higher SNR. The transmit efficiency within the VOI was increased using the dielectric pad, but since the RF pulses were calibrated separately without and with the pad, this did not impact the SNR: the slight increase in RF pulse length without the pad is irrelevant for DWS as TE is long (105 ms) and kept constant. However, for applications with short echo times (e.g., point-resolved MRS, TE~30 ms), the increased transmit efficiency due to the pad would improve SNR since TE can be further minimized, along with the added benefit of a reduced chemical shift displacement³.

For SNR-challenged local MR applications such as high b-factor DWS, high-permittivity materials significantly improve B₁^- sensitivity, allowing for robust in vivo measurements at ultrahigh b-factors in ultrafield systems that push the limits of current hardware capabilities.