In a just published paper in PNAS by Le Bihan and colleagues, a technique called diffusion MRI is used to measure the activation of the brain. This is rather unusual. Diffusion MRI is normally used to measure the diffusion, or movement, of water in the brain. Grey matter is relatively disorganized and water is less restricted than in white matter, where myelinated fibres constrain the direction of molecular movements. This is illustrated by a figure that I have made for an upcoming Elsevier textbook (full size image opens in new window).
Using this method it is possible to map out the major fibre tracts in the brain. Even better, it is possible to do tractography: following the white matter fibre bundles from a seed point and calculate the physical connections from this (see figure below)
Le Bihan and colleagues takes diffusion MRI one step further by applying it to study the brain's activation! The background assumption is: as neurons get (relatively more) activated, a lot of physical movement occurs in and around the neurons. The neuron consumes more oxygen and nutrients; internally in the cell, a lot of movement of these substances, and the movement of signal substances occur, inaddition to energy distribution throughout the cell; and the whole cascade of neurotransmitter release and effects is due to movement across the synaptic cleft. So neural firing causes movement both inside and outside the cell.
Now, what did diffusion MRI do again? It measures the movement (!) of water molecules in the brain. As a brain region gets activated, molecules move relatively more than at rest. As a reult, it should be possible to measure brain activity by looking at where in the brain that molecule movement increases. This is exactly what Le Bihan and colleagues have done. And they provide neat results that it actually works! And, as they argue, diffusion fMRI is a more direct measure of neural activation than the more used BOLD fMRI, which is an index of a complex and delayed mechanism of relative blood oxygenation in regions of the brain.
What remains to find out is what the signal really represents. While it is thought that neuroimaning tools such as BOLD fMRI and EEG measures the activation and energy consumption in the dendrites, we know little about the underlying neural mechanism in diffusion fMRI. Could it be mostly due to movement across the synaptic cleft, and hence be a measure of action potentials; could it be due to movement within the cells; or due to transport of molecules (oxygen + nutrients) across the cell membrane; or all at once? Today, this is an open question. But the mere idea of having yet another MRI tool for measuring the brain's activation, and with good spatial resolution plus the promise of better temporal resolution, is really worth noticing.
We'll be tracking the development of this tool closely!
-Thomas
how is this different than perfusion?
Put (very!) simply, diffusion measures the movement of water. Perfusion measures the movement of blood. While diffusion is a continous measurement, a good example is arterial spin labelling, where a magnetic pulse is applied to the blood supplying the brain. This is typically done at the level of the neck. So with ASL you “label” a part of the blood magnetically, and measure how it floods into the brain.
Hope this is clear.
-Thomas
[...] I found this link over BrainEthics really interesting. Just go and read it, it's about a new technique in fMRI, diffusion fMRI. [...]
New brain scan detects ‘instant’ biological changes
Brain Ethics have picked up on a new development in fMRI brain scanning technology that has the potential to detect fast changes in brain activation. Research just published by neuroscientist Denis Le Bihan and his team has found that changes in brain …
[...] This is not as simple as it sounds. Genes simply do not merely encode how a cell is to look like or function. Genes only react to the environment in which they are situated. The development of, say, a hippocampal cell is not encoded in the genome per se; it stems from the influence of the local environment of that part of the brain, and how the brain cell (at the developmental stage actually more like a stem cell) migrates and connects to the network that will develop into the hippocampus. Neuroscience is certainly sorting out the nitty-gritty details on this, with the fantastic work by people such as Pasco Rakic. IMHO, even neuroimaging cannot escape this turn: we must move from the "blobology" of fMRI, PET and all other neuroimaging methods (this even applies to the clever diffusion fMRI, as described recently), and towards a better understanding of what goes on within these blobs. Just as I briefly mentioned in my post about Nikos Logothetis. [...]
[...] Recent advances have led to an increase in the spatio-temporal resolution of techniques such as functional magnetic resonance imaging (fMRI). These advances mean that it will be possible, in the not-too-distant future, to observe the activity of small groups of neurons, or perhaps individual cells, in real time.The novel use of diffusion MRI, brought to my attention by BrainEthics, is described in a recent PNAS paper. [...]