You are currently viewing Groundbreaking imaging reveals the human brain at nanoscale resolution

Groundbreaking imaging reveals the human brain at nanoscale resolution

This is a digital reconstruction of neurons as seen through an electron microscope. Neurons are colored according to their relative size. These neurons, from an extremely small part of the human brain, are instrumental for all of human consciousness: thoughts, feelings, beliefs, creativity, memory and much more.

For something as central to all human existence and experience as the brain, many aspects of it remain completely mysterious. Thanks to revolutionary new digital images, models and 3D maps created by Google Research and scientists at Harvard University, the physical structure of the brain has never been so clear.

The technology in the game is remarkable. The team, consisting of researchers from Google and Harvard University’s Lichtman Lab, combined electron microscopy and cutting-edge artificial intelligence to inspect a tiny brain sample — just one cubic millimeter (0.000002 pint) — and create a litany of new images, models , and cards.

3D visualization of neural networks showing interconnected neurons with dendrites in different colors such as blue, blue and orange on a dark background.
Neurons – the messengers of life.

There are about 57,000 cells and 150 million synapses, the vital connections between neurons, in this single cubic millimeter of brain. The initial incision was made by a 45-year-old woman during a surgical procedure to treat epilepsy, and this small piece of brain was then cut into about 5,000 individual slices, each about 34 nanometers thick, so that they could be imaged individually using electron microscopy. And yes, this “brain delicacy” is quite remarkable in itself and useful for other brain research.

Highly detailed, colorful 3D rendering of neural networks and brain cells, featuring a complex network of interconnected lines in a varied palette on a dark background.
This 3D map shows the incoming axons for a single cell in the brain.

Neuroscientists Viren Jain of Google and Jeff Lichtman of Harvard—the namesake of the university’s Lichtman Lab—worked together, along with dozens of others, to take thousands of microscopic images of brain tissue and reconstruct them using custom AI models into an entire 3D sample. The project took about a decade to complete and includes 1.4 petabytes (1,400 terabytes) of data that is publicly available. Treating the entire human brain the same way would be one billion terabytes, which is almost as much digital data as the world creates in a year.

Detailed 3D rendering of a neuron cell with an extensive network of dendrites and axons highlighted in blue and green, set against a dark background.
The larger white object is a neuron and all the filaments connected to it are synapses.

“It’s a little humbling,” Wein says of the project. “How are we ever going to come to terms with all this complexity?” The average adult brain is about 1,200 centimeters3or 1.2 million millimeters3. The decade-long project to reconstruct part of the brain looked at just 0.000083 percent of a typical adult’s brain.

“The word ‘fragment’ is ironic,” says Lichtman. “One terabyte to most people is gigantic, but a fragment of a human brain – just one tiny, tiny bit of a human brain – is still thousands of terabytes.”

This view shows six layers of the brain at an extremely microscopic level. This is a reduced view of a sample of cerebral cortex as seen through multiple microscopic images.

This type of work constituted Lichtman’s entire illustrious career. He is a specialist in the burgeoning field of “connectomics,” which is like genomics, but for the brain. Lichtman and his team are working to create complete, detailed images of all brain structures at the single-cell level. The goal is, by making this map, to gain critical insight into brain function and related pathology.

3D rendered image of a neuron with detailed dendrites colored in yellow and blue, synapses showing the complex network of neural connections.  The scale bar at the bottom right shows the size.
These images show one of the axon spirals that researchers have discovered but cannot yet explain. In this 3D model, axons (blue) are connected to, and in some cases physically on, the surface of another cell (yellow).

The initial stage of this grand project is already progressing. The team found a small number of axons formed in a strange swirling pattern unlike anything seen before. However, given that the small part of the brain was taken from a woman with epilepsy, it is unclear whether this spiral was a result of her illness.

Electron micrograph showing intricate details of cell structures, with focus on specific cell indicated by red markers.  A scale bar is available for size reference.
Another mysterious structure that Lichtman and others discovered.

Last year, Lichtman called the brain “depressingly complex” in an interview with Journal of Harvard. However, given the progress he and his team have made, he has reason for some optimism.

Microscopic image showing a dense network of neurons with complex branching, highlighted in different bright colors on a dark background, highlighting the complex connectivity of the brain.
A single neuron (white) receives electrical signals that determine whether it will fire or not. These signals arrive via axons. Greens can deliver a “fire” signal, while blues can tell a neuron not to fire. The human brain has about 86 billion of these little white neurons, therefore trillions of connections.

“If we get to a point where working with a whole mouse brain becomes routine, you could consider doing it in, say, animal models of autism.” There is a level of understanding about the brain that does not currently exist. We know about the outward manifestations of behavior. We know of some of the molecules that are disordered. But between them, the wiring diagrams, until now there was no way for me to see them. Now there is a way,” Lichtman explained.


Image Credits: Google Research and Lichtman Lab, Harvard University. Translations by D. Berger. The full data release is available via Google, and the research is detailed in a paper published in Science.

Leave a Reply