Brain CSI scan protocol considerations

Echo Time (TE) considerations

The recommendation is to acquire MRS data at short TE and, time permitting, to include an intermediate TE scan.

• A short TE (e.g., 20 to 40 ms) is useful to demonstrate myoinositol (MI), glutamine/glutamate (Glx) complexes and lipids.
  • These metabolites are used to characterize most neurological diseases, such as tumors, metabolic and neuro-degenerative disorders, seizures, chronic pain syndrome, and disorders of myelination. They are also used in monitoring therapy for these diseases.
  • This is the recommended TE if only one MRS sequence is considered for the examination; however, the choice of TE would also depend on the clinical indication. For example, in the characterization of neuro-degenerative disorders such as Alzheimer’s disease, short TE MRS is recommended to ensure obtaining information on metabolites only detected with short TE MRS, such as myoinositol and the Glx complexes.
• An intermediate TE (e.g., 144 ms) has a number of advantages over a short TE MRS but provides information on fewer metabolites. Typically, an intermediate TE scan is acquired after the short TE scan, time permitting, for the following reasons:
  • To view the lactate doublet, which is inverted relative to the long T2 species such as the creatine, choline, and N-acetyl resonances.
  • To differentiate lactate and alanine from lipids around 1.3 to 1.4 ppm by J-modulation/inversion of the lactate and alanine doublet peaks.
  • For a better-defined baseline and less baseline distortion compared with short TE.
  • There is no artifactual NAA. The peak in the 2.0 to 2.05 range can only be attributed to NAA rather than superimposed Glx complex peaks in the 2.05 to 2.5 ppm range.
  • To view the presence of lipids that may infer more significance than at short TE.
  • For improved reproducibility and accuracy, particularly for quantifying Cho and NAA peaks.
• A long TE (e.,g. 288 ms) has increased metabolite T2 decay that results in less signal from NAA, Cho, and Cr relative to the baseline noise, and hence the signal to noise is lower than at short and intermediate TE measurements.
  • Experienced spectroscopy users acquire a long TE scan for comparison with normative data.
  • The lactate doublet is fully refocused at TE 288 ms, but even the signals from the long T2 species experience substantial decay, and therefore the 288 ms choice is not recommended.

TR considerations

• A TR of 1500 or 2000 ms enables the comparison with spectra acquired with PROBE-P in its single voxel mode.
• Use the same TR so that you can compare spectra acquired on different dates, or between different patients.  

2D spectroscopy considerations

• The 2D CSI reconstruction process creates 256 small chemical shift images that are stored in a standard MR image in a separate series. The 256 images span the frequency range from 4.3 to 0.49 PPM. Spectra and metabolite maps are extracted from the individual CSI images with READY View.

Scan parameters

• Typical examination times range from 3 to 15 minutes.
• Scan selections: MRS Mode, Spectroscopy family, PROBE-P pulse.

Importance of the linewidth value

• The LnWdth value is critical to the success of a spectroscopy scan. The shim process (part of auto prescan) attempts to improve the homogeneity through the voxel by adjusting the gradient currents. The linewidth value is a measure of the voxel homogeneity; the smaller the linewidth, the better the homogeneity. For 2D CSI acquisitions, a linewidth value less than 0.2 ppm is acceptable.
  • For 1.5T, 13Hz

    Remember however, that the linewidth usually increases as the CSI phase encoding volume increases. If the linewidth is greater than 0.2 ppm, check the position of the voxel in the Graphic Rx viewport and, if necessary, reposition the voxel to avoid magnetically in-homogeneous regions of the anatomy. Repeat Auto Prescan with the repositioned voxel. When you are satisfied with the Auto Prescan results, proceed to scan.

• The 2D CSI SNR depends on TE, TR, NEX, Frequency, Phase, and Voxel Thickness.
• One cubic centimeter (1 cc), nominal CSI volumes provide reasonable SNR and spatial resolution with the head coil. If the FOV value = Frequency value = Phase value, each edge of the CSI pixel is 1 cm in length. A 1 cm voxel thickness gives a volume of 1 cc.
• The FOV divided by Frequency/Phase determines the lengths of the in-plane edges of the CSI voxel. The length of the third edge is the same as the Voxel Thickness of the PRESS VOI.

3D Spectroscopy considerations

• Scan selections: MRS Mode, Spectroscopy family, PROBE-P pulse.
• The 3D CSI reconstruction process creates 256 small chemical shift images that are stored as a standard MR image. The complex CSI storage images, two for each slice - i.e., real and imaginary - are stored in a separate series. The CSI images span the frequency range from 4.3 to 0.49 PPM. Spectra and metabolite maps for each slice along the third dimension are extracted from the individual CSI images with READY View.
• Practical examination times range from 9 to 35 minutes.
• The 3D CSI SNR depends on TE, TR, NEX, Frequency, Phase, and # of CSI Slices and CSI Slice Thickness.

Importance of the linewidth value

• The linewidth value is critical to the success of a spectroscopy scan. The shim process (part of Auto PreScan) attempts to improve the homogeneity through the voxel by adjusting the gradient currents. The LnWdth value is a measure of the voxel homogeneity; the smaller the linewidth, the better the homogeneity. For 3D CSI, a linewidth value less than 0.25ppm is acceptable.
  • For 1.5T 16Hz

    If the linewidth is greater than 0.25 ppm, check the position of the voxel in the Graphic Rx viewport and, if necessary, reposition the voxel to avoid magnetically inhomogeneous regions of the anatomy. Repeat Auto Prescan with the repositioned voxel. When you are satisfied with the Auto Prescan results, proceed to scan.