The brain is made of up billions of neurons and even more connections between neurons. We can get an idea of how neurons are connected across the human brain using a type of brain scan called diffusion-weighted magnetic resonance imaging (dMRI).
How does dMRI work? It actually measures the direction that water moves throughout the brain. Sometimes, water can move in all directions equally easily – like how one water molecule in a glass of water can go anywhere else in the glass. Other times, water’s motion is constrained by its location, like how a spilled glass of water spreads out across a table but not through the table.
The connections between neurons are made by neurons’ axons, which often travel together in bundles, sort of like a neural highway. These bundles constrain the movement of water in the brain so that water tends to movealong the bundle’s trajectory, not through the bundle itself. Since dMRI measures water flow in the brain, we can use it to infer the orientation of neural fiber bundles. (Individual axons are unfortunately too small to measure directly using dMRI).
Complex computational algorithms can take in the orientation of fiber bundles at every point in the brain and reconstruct the tractography of the brain.
This whole-brain tractography was created with a method called constrained spherical deconvolution using dMRI data from a high-sensitivity MRI (the Connectome scanner) at the Martinos Center for Brain Imaging, Massachusetts General Hospital, in Charlestown, Boston, Massachusetts.
Contributed by Kevin Sitek, a 4th-year Harvard PhD student in Speech and Hearing Bioscience and Technology doing research at the MIT McGovern Institute for Brain Research.