GOOGLE AND HARVARD TEAMS UP TO CREATE THE MOST DETAILED MAP OF THE HUMAN BRAIN

One of the most important and complicated objects in the world is the human brain. However, scientists still don't fully understand the three pounds of tissue that govern every decision, emotion, and reflex in our lives. A recent collaboration between the research teams at Google and Harvard has successfully extracted an amazing amount of data from a mere 3 mm sample of brain tissue. The scientists used machine learning modelling to produce the highest-resolution brain tissue map ever made. Researchers can see 150 million synapses—connections between neurons—and 57,000 cells on this brain map. 

Despite the fact that the brain tissue was only half the size of a rice grain, the study's raw data could have stored 14,000 full-length films or 2,800 laptops' worth of data. As a neuroscientist at Google and co-author of the study, Viren Jain told Nature News, "It's a little bit humbling." "How will we ever truly understand all of this complexity?"

To help people understand some of the intricacies of the brain, the researchers have made their 1.4 petabytes of raw data publicly available for use and review.

The 45-year-old patient whose brain tissue was used had surgery to treat her epilepsy. Since brain biopsies are uncommon and typically only involve tumours, this was a unique example of living brain tissue that the researchers were able to preserve in resin. Furthermore, because brains decompose quickly, cadavers—which are frequently used in medical research—are not useful.

The tissue had to be divided into five thousand slices, each measuring thirty nanometers in thickness after the scientists had preserved it. After that, the slices were inspected using an electron microscope designed especially for this research. After more than a year, artificial intelligence took over to ensure that every neuron had the proper synapses connected to it when reconstructing the images.

Surprising results have already been obtained from the 3D reconstruction, which includes all tissue elements, including glial cells, blood vessels, and the myelin sheath surrounding the neuron. According to Jeff Lichtman, the head of the Harvard lab working on the project, "There were just so many things in it that were incompatible with what you would read in a textbook." Lichtman is a molecular and cell biologist.

The fact that numerous neurons are connected to numerous synapses—in one instance, 50.96% of neurons are connected to just one synapse, and 99% of neurons have fewer than three synapse connections—is one of the most fascinating findings. Although the exact meaning of these extra-connected neurons is unknown to scientists, the current theory suggests that this may be the appearance of a well-learned response requiring little thought. For example, skilled drivers don't even realize they are braking with their feet because it is so second nature to them.

The symmetrical dendrites of pyramidal neurons—the branches that carry information away from a neuron—is one of several other facts that astounded scientists.

Additionally, they discovered previously unseen whorls of axons—branchlets that transmit information from synapses to the cell body—circling around themselves. It is easy to see how many more avenues for neuron research will become available with more eyes carefully going over the data.

There have previously been brain maps made, dating back to 1986 when 302 neurons in a roundworm were mapped. Although more intricate maps of other animals have gradually been produced, years and significant technological advancements will be needed before the human brain is fully mapped at this resolution. Up until that point, the group has worked on mapping the hippocampus of a mouse, a brain that is structurally and neuronally quite similar to that of a human.