Embedded Files
A brief novice primer into edited MRS:
The fact that we can make images of the brain or brain structures, or measure brain activity, with MRI, is based on the fact that hydrogen-protons align when placed in a magnetic field. One can change the magnetisation of these protons by applying 'radio-frequency' pulses, which result in these protons precessing in a magnetic field back to equilibrium. Aspects of this precession can be measured.
As our body is full of hydrogen, mainly in the form of water, it is an easy and non-invasive contrast that can be used to visualise structure and activity. Most of the contrast we see in anatomical images is based on the different ways water behaves in different tissues. You can measure for instance how long it takes for magnetisation to get back to equilibrium after an RF pulse. This time to equilibrium is called longitudinal relaxation or; T1). 'Free' water, such as in the cerobrospinal fluid (CSF), takes a long time to get back to equilibrium and has a long T1, whereas bound water, such as in fatty tissue, has a short T1. So if the T1 is what you're interested in and you measure very quickly after the RF pulse, CSF will appear dark as it has hardly any signal yet, whereas fat appears bright because more magnetisation has returned to equilibrium.
What about MRS?
Water is the most abundant molecule in the brain. However, if the strong water signal is suppressed, a collection of other metabolites (with much lower concentration) can be detected. Due to their chemical environment, which is mainly determined by electron density, hydrogen spins of different metabolites precess with slightly different frequencies. An example spectrum, acquired from a single-voxel in the brain, is shown below, with different peaks along the spectrum, most prominently NAA, Choline and Creatine.
And what about GABA?
The concentration of GABA in the brain is a couple of factors less than other metabolites shown below, but the signal of GABA is very similar to that of Creatine. The GABA signal is hidden underneath these larger peaks and therefore it is very difficult to measure GABA from a spectrum like this. Luckily there are some things we can do about that!
GABA actually consists of three separate signals corresponding to three different CH2 groups; (as can be seen in the figure above), each at a different location along the spectrum because each group has a slightly different chemical environment. However, these different groups are coupled to each other, so if you modulate one group, it has an effect on the other group. And we use this coupling to measure GABA. By acquiring a spectrum as shown above, but applying an additional RF pulse to GABA at 1.9ppm (called 'editing') in half of the scan, and not in the other half, we also modulate the GABA signal at 3 ppm (see below), because GABA at 1.9 and 3 ppm are coupled to each other. By subtracting the 'edited' part from the unedited part, we end up with a spectrum that only contains the signals affected by the GABA-editing pulse (see below). Creatine is not affected and will be subtracted out, but GABA is affected and therefore appears in our edited spectrum. We can quantify this peak and there we have our GABA concentration! Interestingly, we also obtain measures of Glx (Glutamate + Glutamine) as indirect markers of excitation and metabolism!
Back in the day, I wrote a review on the use and application of GABA-MRS, which you can find here. New and better papers are being published all the time though!
For a paper that we wrote with a number of people, on the practical use and analysis of MEGA-PRESS, click here
- Puts, N.A., Edden, R.A., (2012) In vivo magnetic resonance spectroscopy of GABA: A methodological review, Progress in Nuclear Magnetic Resonance Spectroscopy, 60:29-41. PDF
- Mullins, P.G., McGonigle, D.J., O’Gorman, R., Puts, N.A., Vidyasagar, R., 1st Cardiff Symposium on MRS of GABA, Evans, C.J., Edden, R.A.E., (2012). “Current Practice in the use of MEGA-PRESS spectroscopy for the detection of GABA”, Neuroimage. doi:pii: S1053-8119(12)01177-9. 10.1016/j.neuroimage.2012.12.004. PDF
Google Sites
Report abuse