Synopsis

Keywords: Spectroscopy, Brain

The prefrontal cortex plays a central role in controlling executive functions such as decision-making and emotional-monitoring and has thus been the subject of several ¹H-MRS studies on psychiatric illnesses. This region, however, is sensitive to motion-induced increases in linewidths due to the voxel moving away from the originally shimmed location. Here, we combined metabolite cycling (MC) with continuous prospective frequency correction to better account for frequency instabilities when performing single-voxel brain MRS at 7T. We find that MC sLASER with continuous prospective frequency correction reduces frequency instability, water linewidths, and N-acetylaspartate linewidths relative to semi-LASER acquisitions without prospective frequency correction.

Purpose

The prefrontal cortex (PFC) plays a central role in controlling executive functions such as decision-making (1) and has thus been the subject of ¹H-MRS studies on psychiatric illnesses (2). This region is more challenging to measure from, however, due to its proximity to air-tissue interfaces like the sinuses which causes significant B₀ inhomogeneity from susceptibility effects. Although the region can be shimmed to improve B₀ homogeneity, B₀ magnetic field conditions rapidly deteriorate in areas adjacent to the originally shimmed location (3, 4). Thus, head and breathing motion can result in linewidth increases due to the voxel moving away from the originally shimmed location. We have previously demonstrated that continuous prospective frequency correction with metabolite cycling (MC) decreases macromolecule (MM) contamination and subtraction artifacts caused by frequency instabilities when performing MM-suppressed GABA-editing in the medial PFC relative to that with interleaved prospective frequency correction (4). Here, we combined MC with continuous prospective frequency correction to better account for frequency instabilities caused by motion and gradient-induced heating/cooling when performing semi-LASER and compare its performance to water-suppressed semi-LASER without prospective frequency correction at 7T.

Methods

Figure 1 shows the MC sLASER sequence which includes an adiabatic MC pulse in place of water suppression pre-pulses. These pulses invert the upfield and downfield resonances in alternate transients while still retaining the full water signal. The high-intensity water signal allows for prospective frequency correction at each transient by Fourier transforming each FID, calculating the frequency offset from the nominal water frequency, and adding it to the spectrometer center frequency prior to the next transient acquisition.

Acquisitions were performed in 4 healthy volunteers (2 female, age 35.5±15.7 years) on a 7T Philips Achieva scanner in a 2.5x2x2 cm³ voxel in the left dorsolateral PFC (dlPFC) with TE=30 ms, spectral width=4 kHz, and 32-channel receive head coil (Nova Medical). Additional acquisition parameters for the MC sLASER scans include TR=6.7 s and 96 transients (total scan time=11 mins). The transmit frequency of the MC pre-inversion pulse was shifted by -260 Hz and 260 Hz for the upfield and downfield RF inversion, respectively. Additional acquisition parameters for the water-suppressed sLASER scans (WS sLASER) include VAPOR water suppression, TR=4.5 s and 128 transients (total scan time=10 min). Water-unsuppressed sLASER (WU sLASER) scans were acquired with 32 transients and the same parameters as the water-suppressed scans except for the water suppression.

Data were analyzed using in-house software and Gannet (5) with time domain retrospective phase and frequency correction (6). The difference in the water frequency (ΔF₀) from the nominal water frequency (4.68 ppm) and water linewidths were calculated from the water-unsuppressed sLASER scans. Signal-to-noise ratios (SNR) were calculated as the N-acetylaspartate (NAA) peak divided by the root-mean-square of the noise, while the linewidths of the metabolite spectra were calculated from the NAA peak. Paired two-sided t-tests were performed to assess group differences, and a p-value <0.05 was chosen to be statistically significant.

Results

Figure 2a shows an example left dlPFC voxel placement. Figure 2b shows example MC sLASER and WS sLASER spectra from one volunteer showing 1.4 Hz narrower linewidths in the MC sLASER spectrum than in the WS sLASER spectrum. The change in the water frequency and water linewidths over time as measured from the unsuppressed water signal with prospective frequency correction (MC sLASER) and without prospective frequency correction (WU sLASER) are shown in figure 3. MC sLASER shows minimal change in the water frequency over time while WU sLASER shows significantly greater change in the water frequency (up to 12 Hz) over time. MC sLASER also shows significantly lower water linewidths over time than WU sLASER. Figure 4 shows spectral quality metrics for MC sLASER, WU sLASER, and WS sLASER across subjects. The average ΔF₀ and water linewidths are 9 Hz (p=0.013) and 1.2 Hz lower (p=0.011), respectively, in MC sLASER than in WU sLASER. Although SNR is equivalent between MC sLASER and WS sLASER at ~815, NAA linewidths are lower in MC sLASER than in WS sLASER (p=0.017) with a median linewidth reduction of 1.4 Hz across subjects, similar to the reduction in the water linewidths.

Discussion

It can be seen that MC sLASER with continuous prospective frequency correction decreases the amount of B₀ field instability relative to conventional WS sLASER without prospective frequency correction. MC sLASER is also shown to reduce linewidths relative to WU sLASER and WS sLASER even after retrospective frequency correction. The reduced linewidths with MC sLASER are likely due to a reduction in voxel movement away from the originally shimmed location. Conversely, the larger linewidths with WS sLASER relative to MC sLASER are likely due to the voxel moving into adjacent regions which contain greater B₀ inhomogeneity which is further amplified by the ultra-high field strength of 7T. Although demonstrated here for the dlPFC, MC with prospective frequency correction is likely to provide the same benefit to other difficult-to-shim regions such as the temporal lobes. Larger reductions in linewidths with MC sLASER relative to WS sLASER are likely to be seen in pediatric and clinical populations which are more prone to movement than the healthy adult participants studied here and after gradient-intensive scans (e.g. fMRI, DTI scans).

Acknowledgements

This work was funded by the Cancer Prevention and Research Institute of Texas (CPRIT) (Grant Number: RR180056, Principal Investigator: Anke Henning, PhD) and by the Center for Depression Research and Clinical Care (Principal Investigator: Madhukar H. Trivedi, MD), Hersh Foundation.