MR spectroscopy

MRI · reads

Proton MR spectroscopy (¹H MRS) is a noninvasive MRI-based method that estimates tissue metabolite profiles from proton signals, which are separated by their chemical shift. It chiefly detects NAA, creatine and choline, and sometimes lactate, myo-inositol and Glx (the combined glutamate and glutamine signal), and the signal intensities can be scaled into concentration estimates. Sequences of radiofrequency pulses localize a small tissue volume, and a difference-editing approach is commonly used to separate overlapping signals and quantify GABA.

Spatial resolution
3 mm (point spread)
Temporal resolution
2 min (sample interval)
Field of view
1.23 L (acquired volume)
Reads
metabolite concentration (NAA); metabolite concentration (creatine); metabolite concentration (choline); metabolite concentration (lactate); metabolite concentration (myo-inositol); metabolite concentration (Glx); metabolite concentration (GABA)
Destructive?
non-destructive
Prerequisites
none beyond ordinary participation
Theoretical field of view
1.23 L (reachable envelope)

Description

Proton MR spectroscopy (¹H MRS) is a noninvasive MRI-based method that estimates tissue metabolite profiles from proton signals separated by chemical shift, measured in parts per million (ppm). It chiefly detects NAA, creatine and choline, and sometimes lactate, myo-inositol and Glx (the combined glutamate and glutamine signal). The unitless signal intensities can be scaled to concentration estimates. Single-voxel STEAM (three 90° pulses, stimulated echo) and PRESS (90° excitation plus two refocusing pulses, spin echo) localize the intersection of three slices. MRSI is a hybrid spectroscopy-imaging experiment and typically uses smaller voxels with longer scans. MEGA-PRESS difference editing exploits known GABA couplings to separate its signal from stronger overlapping metabolite signals and quantify GABA, commonly at 3 T. Relative to 1.5 T, 3 T improves signal-to-noise ratio, reduces spectral overlap and makes metabolite measurement more reliable, but it also increases the chemical shift displacement error, a mismatch in where different resonances are localized. With conventional small-bandwidth refocusing pulses this displacement can exceed 30% of the voxel dimension across the 3 ppm range in one direction, even at 3 T. For applications that look for subtler metabolite changes, these conventional protocols are severely limited at 3 T and at ultra-high fields of 7 T and above.

Download the dataset (JSON, described by its JSON Schema).

Loading…