Researchers at Baylor College of Medicine (BCM), the Allen Institute, and Princeton University have created the largest 3D wiring diagram of a mouse brain — which is extremely similar to our own — in the hopes of learning principles from the brain that might aid progress AI, according to a press release.
The map you see above depicts the intricate structures and connections of roughly 200,000 brain cells and almost 500 million synapses — all hidden in a cubic millimeter chunk of mouse brain the size of a grain of sand.
What's more, the dataset is unique in that it includes recordings of activity patterns elicited by a variety of complex visual stimuli, ranging from YouTube clips to Hollywood movies, for approximately 75,000 brain cells in the same brain volume that was used to generate the connectivity map.
The map and data set are now available to the public.
500 million synapses’ worth of data available
The map took five years to create, with three distinct stages. The researchers focused their investigation on the visual neocortex, a brain area that is critical for visual perception. The researchers collected measurements of the mouse's typical brain activity while it was still alive in the first stage, producing over 70,000 pictures of active brain cells.
Then, scientists cut out a small piece of the brain and sliced it into more than 25,000 ultra-thin pieces, and took more than 150 million high-resolution images of those pieces using electron microscopy.
The research is part of the Machine Intelligence from Cortical Networks (MICrONS) program, which aims to improve machine-learning algorithms and AI by reverse-engineering the cerebral cortex, which is responsible for higher functions such as planning and reasoning in mammals.
The map will be "invaluable" for neuroscientists trying to figure out how the brain processes information along with neocortical circuits, as well as researchers trying to figure out how to treat brain illnesses involving faulty wiring or connections, according to the researchers.
"Computer programs have gone through a massive revolution in the past decade, but something is still missing: today’s programs are easy to fool and they generalize poorly to new situations. In many ways they are much dumber than a mouse," said Dr. Xaq Pitkow, associate professor of neuroscience and one of the lead scientists of MICrONS. "Our new exquisitely detailed data about the mouse brain gives us a chance to find what makes the biological brain smarter. And that could help us develop smarter computers."