Advanced Computing in the Age of AI | Sunday, May 26, 2024

Harvard/Google Use AI to Help Produce Astonishing 3D Map of Brain Tissue 

Although LLMs are getting all the notice lately, AI techniques of many varieties are being infused throughout science. For example, Harvard researchers, Google, and colleagues published a 3D map in Science this week that reveals a small chunk of as human brain in astonishing detail. Imaging the roughly cubic millimeter of tissue produced 1.4 petabytes of data.

Nature has a nice summary article and quotes Viren Jain, a neuroscientist at Google in Mountain View, California, and a co-author of the paper, saying “It’s a little bit humbling. How are we ever going to really come to terms with all this complexity?” Jain’s team then built artificial-intelligence models that were able to stitch the microscope images together to reconstruct the whole sample in 3D. “I remember this moment, going into the map and looking at one individual synapse from this woman’s brain, and then zooming out into these other millions of pixels,” says Jain. “It felt sort of spiritual.”

The work itself has been going on for roughly ten years. “About 10 years ago, a small piece of human brain arrived in the lab of Dr. Jeffrey Lichtman at Harvard. It came directly from an operating room of a nearby hospital, where it was excised from an epilepsy patient undergoing a procedure to reduce her seizures, reported the Boston Globe.

A single neuron (white) shown with 5,600 of the axons (blue) that connect to it. The synapses that make these connections are shown in green.Credit: Google Research & Lichtman Lab (Harvard University). Renderings by D. Berger (Harvard University)

“In the years that followed, Lichtman’s team methodically reconstructed the byzantine wiring patterns of the brain by feeding the 1-cubic-millimeter sample into a $6 million device that sliced it into impossibly thin slivers. Then, using images of those slivers taken by electron microscopy, they painstakingly recreated the intricate latticework connecting individual cells to one another.”

The images are remarkable. The abstract give a sense of the size of the project:

“Presented here is a computationally intensive reconstruction of the ultrastructure of a cubic millimeter of human temporal cortex that was surgically removed to gain access to an underlying epileptic focus. It contains about 57,000 cells, about 230 millimeters of blood vessels, and about 150 million synapses and comprises 1.4 petabytes. Our analysis showed that glia outnumber neurons 2:1, oligodendrocytes were the most common cell, deep layer excitatory neurons could be classified on the basis of dendritic orientation, and among thousands of weak connections to each neuron, there exist rare powerful axonal inputs of up to 50 synapses. Further studies using this resource may bring valuable insights into the mysteries of the human brain.”

The full paper is available at Science, but Google and the researchers are providing public access to the dataset and Neurglancer viewer to look through the data.

Link to Science paper (A petavoxel fragment of human cerebral cortex reconstructed at nanoscale resolution), https://www.science.org/doi/10.1126/science.adk4858

Link to Boston Globe Article, written by Adam Piore, https://www.bostonglobe.com/2024/05/10/metro/google-harvard-brain-map-images/

Link to Nature article, https://www.nature.com/articles/d41586-024-01387-9

EnterpriseAI