The Brain’s Building Blocks: Of Protons and Voxels

What’s in a brain? That which we call a voxel by any other name would sound far less confusing.

Imagine all of the atoms in your brain. Now imagine how these atoms might behave inside of a giant magnet. Fortunately, there’s a technique that makes this thought experiment possible. Structural magnetic resonance imaging (MRI) uses the behavior of basic building blocks—like atoms and their protons within—to form a truly interdisciplinary technique that spans physics, biology, and computer science in order to “photograph” the brain.

Structural MRI is how we know that the frontal cortex grows during development, and that the temporal lobe degenerates in Alzheimer’s disease. Developing or diseased, the brain, like all other living matter, is composed of nature’s most basic building blocks. Of these blocks, protons are uniquely important for MRI. Though many a neuroscience finding has relied upon the generation of bright, beautiful brain scans, we less often discuss how scans result from our ability to measure the brain’s building blocks. How does neuroimaging leverage the brain’s quantum properties to acquire data? How does basic physics help us peer into an individual’s neural structure?

“Monochromatic signals appear; the story of a living brain emerges.”

The answer: A little bit of water, and a lot of something called a voxel.

An MRI machine is, per its moniker, a giant magnet. Most human research machines range from 1.5 – 3 Tesla (the unit for magnetic field strength)—this is very, very strong. This level of strong can pull up to 2000 pounds, roughly the weight of a small car. Unlike cars, the brain isn’t made of metal but of water. We can think of the brain as bulk matter made of water molecules, whose nuclei contain hydrogen protons that hold the potential for magnetic properties akin to a tiny magnet. The phenomenon responsible for this magnetic property is called spin: Spin is an intrinsic movement for a subatomic particle, like a proton. Basic physics tells us that movement creates current, and current creates a magnetic field. Inventors of MRI recognized this law. Critically, there is just enough water in the brain—and thus, the protons in this water—to create a magnetic moment that can be contrasted against even stronger magnetic forces. Thus, in an MRI machine, the brain’s collective protons form a net magnetic moment due to spin, providing researchers an opportunity to probe the brain using little more than electromagnetic forces.

Spin, and thus its magnetic properties, can change. When a living brain enters the bore of an MRI machine, laws of quantum mechanics dictate the behavior of the brain’s protons the moment they find themselves immersed in a magnetic field. By physical law, when protons experience a magnetic field, their spin must abide with the direction of magnetism. The moment a body enters the MRI machine, the protons within the water of the brain are inclined to align with the magnet’s field. But other forces, like radio-frequency waves, can tip these protons out of sync with the magnet. The protons tip into a new direction, no longer in their original alignment. This is exactly what neuroimaging does: Researchers prompt the machine to emit radio-frequency pulses at specific locations in the brain, where interactions between protons and the magnet begin the process of building brain scans.

A depiction of how MRI machines take advantage of proton spin properties to gather images of the brain. Illustration by Huixuan Liang.

Depending upon whether a proton lies within gray matter, white matter, or cerebrospinal fluid, its spin is disrupted in a unique way, causing its return to baseline to be distinct from that of nearby brain tissue. For example, protons in neuronal tissue will return to their initial spin at a slightly different time scale than those in glia. The scanner notices this discrepancy and encodes this as a signal. In brain scans, this signal is relayed through black and white imagery. Monochromatic signals appear; the story of a living brain emerges.

“The mighty little voxel carries researchers from broad visual inspection to small but computationally powerful measurements of the brain.”

But this is only the beginning, and building the story of a brain requires much more work. An individual’s neural structure is, in its own right, fascinating. Yet the roles of research require many scans from many individuals, with multiple statistical tests applied to each scan. Re-enter the voxel: A portmanteau of “volume” and “pixel,” the voxel is a 3-dimensional unit that embeds the signals in brain scans. As the MRI machine scans through each dimension of the brain millimeter by millimeter, voxels are formed to enclose the signals created by protons-magnet interactions. The brain as scanned is made of numerous voxels, each of which carries indirect measurements about the density, shape, or size of neurons (or white matter!). Perhaps more surprising, the amount of water required to produce these measurements is small in comparison to the number of voxels yielded by just one scan.

But what do voxels tell us about brain structure? A voxel by itself is meaningless. This tiny little cube, when compared to those of different individuals, can tell us many stories: How is gray matter different across the lifespan? Do taxi-cab drivers have larger hippocampi than non-taxi-cab drivers? (The answer: Maybe so.) Across many brains, patterns of voxels become meaningful and allow scientists to test research questions with careful interpretation. The mighty little voxel carries researchers from broad visual inspection to small but computationally powerful measurements of the brain.

Counting the brain voxel-by-voxel could prove impossible. That is, the sheer number of voxels in brain images across subjects requires the scientist to relinquish some of the effort to a computer. Techniques such as voxel-based morphometry (VBM) use automated computer processes to transform voxels into meaningful units. In this most-popular neuroimaging method, a voxel undergoes a long journey: Its dimensions are warped into a standard brain template, its components are classified by tissue type, and the even-smaller blocks within the voxel itself are estimated based on the composition of its neighbors. The output of this process takes hours or days, but the result—a map of the brain—is priceless.

After detecting the interactions between protons in the brain and the magnet of the MRI, these signals can be processed to produce a map of the brain. Illustration by Huixuan Liang.

Sometimes, the voxel is controversial. Some studies have argued that differences in how researchers apply the voxel-based method can lead to discrepancies in our understanding of fundamental issues, such as brain structure in autism. Voxel-based neuroimaging methods are also sometimes unsuitable in studies where brain structure is significantly altered, as in the case of lesions due to stroke, and modified steps are necessary to make voxels useful in analyses of these brains.

The question of “what’s in a brain” is not quite answerable at the cellular level by current neuroimaging methods. The voxel remains an indirect measure of the tissues and structures neuroscience wishes to understand. However, that which we call a voxel inches our understanding of the brain forward.

To see Knowing Neurons’ take on how MRI contributes to neuroscience research, look here!


A blue and grey pixelated image of the brain as a symbolic representation of voxels, 3D pixels generated by neuroimaging techniques. Illustrated by Jooyeun Lee.
Feature image illustrated by Jooyeun Lee.


Baron, J.C., Cheletal, G., Desgranges, B., Perchey, G., Landeau, B., de la Sayette, V., & Eustache, F. (2001). In Vivo Mapping of Gray Matter Loss with Voxel-Based Morphometry in Mild Alzheimer’s Disease. Neuroimage.

Geva, S., Baron, J.C., Jones, P.S., Price, C.J., & Warburton, E.A. (2012). A comparison of VLSM and VBM in a cohort of patients with post-stroke aphasia. Neuroimage Clinical.

Katuwal, K.J., Baum, S.A., Cahill, N.D., Dougherty, C.C., Evans, E., Evans, D.W., Moore, G.J., & Michael, A.M. (2016). Inter-Method Discrepancies in Brain Volume Estimation May Drive Inconsistent Findings in Autism. Frontiers in Neuroscience, Brain Imaging Methods.

Lerch, J.P., van der Kouwe, A.J.W., Raznahan, A., Paus, T., Johansen-Berg, H., Miller, K.L., Smith, S.M.,Fischl, B., & Sotiropoulos, S.N. (2017). Studying neuroanatomy using MRI. Nature Neuroscience.

Maguire, E.A., Gadian, D.G., Johnsrude, I.S., Good, Catriona, C.D., Ashburner, J., Frackowiak, R.S.J., & Frith, C.D. (2000). Navigation-related structural change in the hippocampi of taxi drivers. PNAS.

Mechelli, A., Price, C.J., Friston, K.J., & Ashburner, J. (2005). Voxel-Based Morphometry of the Human Brain: Methods and Applications. Current Medical Imaging Review.

Sowell, E.R., Thompson, P.M., Holmes, C.J., Jernigan, T.L., & Toga, A.W. (1999). In vivo evidence for post-adolescent brain maturation in frontal and striatal regions. Nature Neuroscience.

Gabrielle Torre

Gabrielle-Ann is a PhD student at Georgetown University and studies the neural correlates of reading, IQ, and socioeconomic status. She is broadly interested in using neuroimaging methods to ask questions about human cognitive behaviors and abilities. Previously, she studied brain-behavior relationships in healthy aging at the University of Arizona, where she developed a love for literature and creative writing. She still enjoys reading and writing, as well as live music, gender studies, and eating.