Generate fMRI from neurotransmitter consumption

Question

As mentioned in this answer, it's possible to generate an fMRI BOLD signal from neurotransmitter consumption. What equation would be appropriate for this use?

Answer 1

The simplest equation for getting a BOLD signal from neurotransmitter that I could find was in "Tracing Problem Solving in Real Time: fMRI Analysis of the Subject-paced Tower of Hanoi", which itself references many other publications where it was used:

$$H(t)= m \times(t/s)^a\times e^{-(t/s)}$$

The parameters $s$, $a$ and $m$ don't have an explicit meaning. Heuristically, from the text:

$m$ is the magnitude of the response and $s$ is a time scale. The function peaks at time $a \times s$. The parameter a determines the shape of the function such that the larger $a$ is the more narrowly the function will be distributed around its peak.

So it seems you have to fit it to some previous data in the area you're trying to generate data from before it can be used.

Answer 2

It is not clear what neurotransmitter you are interested in, but one model that relates the PET ligand affinity to the cerebral blood volume (CBV) from fMRI looks like this:

_{Source: Neurovascular coupling to D2/D3 dopamine receptor occupancy using simultaneous PET/functional MRI}

You can find several studies, which try to relate neurotransmitters with oxygen consumption. Google made the following suggestions to me:

• Brain hemodynamic changes mediated by dopamine receptors: Role of the cerebral microvasculature in dopamine-mediated neurovascular coupling
• GABA concentrations in the human anterior cingulate cortex predict negative BOLD responses in fMRI

In addition to PET, of special interest is the technique MRS using the same type of scanner that is used for fMRI:

• In vivo magnetic resonance spectroscopy of GABA: A methodological review