Deconstruction of billions of brain circuits

  When Bill Newsome, a neurobiologist at Stanford University School of medicine in California, received a call from Francis Collins, Dean of the National Institutes of health, his first reaction was panic Collins suddenly contacted him and asked Newsome if he would co chair a 10-year brain research project   For Newsome, it sounds like a thankless, unorganized, cumbersome job that can ruin a good summer mood However, after 24 hours of ideological struggle, Newsome's enthusiasm for work eventually defeated panic He accepted the task "It's a great time," Newsome said In the 21st century, the study of the brain is an interesting topic for the academic community "   As early as April 2, US President Barack Obama announced plans to invest $100 million to launch the "brain project" - a project designed to study the brain that could ultimately cost 10 times the budget Similarly, the European Commission has similar ambitions At the beginning of this year, the European Commission announced the launch of its flagship project, human brain engineering, with a budget of 54 million euros ($69 million) in 2013 and a funding of 1 billion euros over the next 10 years   Although the goals of brain projects in the United States and the European Union are different, in fact, they all point to the ultimate problem of neuroscientists: how to organize billions of neurons and trillions of connections or synapses in the human brain to form an effective neural circuit so that human beings can fall in love, go to the battlefield, solve mathematical theorems or write poems In addition, the researchers want to know how the neuronal circuits in the brain change   To achieve this goal, new technologies (such as nanotechnology, genetics, optics, etc.) are needed to be able to capture the EEG activity between neurons, then stimulate these neurons to understand their working state, and draw detailed potential structure circuits "The amount of data generated by the human brain in 30 seconds is equivalent to the amount of information generated by the Hubble Space Telescope during its lifetime," said Konrad kording, a neuroscientist at Northwestern University in Chicago   Researchers are already looking for some entry points to solve the problem In the past few years, people have witnessed amazing technological progress, for example, scientists have built extremely accurate human brain anatomy map Up to now, most neuroscientists are using simpler species, such as mice or worms, to study some basic principles Here, nature looks at some of the technological advances that have helped people better understand how the brain works   Measuring neurons   If researchers want to understand the electronic signals flowing through the brain circuits, they need to record as many neurons as possible simultaneously   At present, researchers mainly measure the activity of neurons by embedding metal electrodes in the brain But this approach is accompanied by great challenges Each electrode needs wires to measure analog signals - voltage changes, but in the process of transmission, signals are easily lost or distorted   In addition, wires must be as thin as hair to avoid tissue damage The progress of electrode technology makes it possible to record hundreds of neurons at the same time But to solve the ultimate puzzle, researchers need to detect as many cells as possible and capture better quality signals   The advent of a new generation of neural exploration devices made of silicon, featuring extreme miniaturization, made it possible In February, the prototype of the device (only 1 cm long, as thin as a dollar bill) was unveiled at the International Solid State Circuit Conference in San Francisco, California The device was launched by IMEC (based in leffin, Belgium), an organization specializing in nanoelectronics   When the device is embedded in the mouse's brain, the electrodes on the IMEC device can simultaneously record information on all cerebral cortex of the mouse This helps neuroscientists break down complex circuits that are interconnected in the brain "This prototype can be expanded," said Peter peumans, director of Bioelectronics at IMEC Within three years, the neurosearch device will have as many as 2000 electrodes and more than 200 wires   Draw a map   Researchers have been collecting information about neuronal activity and circuits, which is necessary for a reliable and highly detailed brain anatomy   For more than a century, various methods of mapping neuroanatomy have been trying to cut the brain into as thin slices as possible, and then scientists will study these slices carefully under the light microscope However, it is very challenging to arrange these astonishing numbers of slices in order to reconstruct the intricate neural network in human brain   Nevertheless, Katrin Amunts and colleagues at the Ulrich research center in Germany announced that they had completed the difficult work last month   By slicing and analyzing a 65 year old woman's brain sample, the researchers have produced the most detailed three-dimensional brain map to date The brain map is a compilation of 7400 brain slices, each 20 microns thick, which are imaged by an optical microscope Two supercomputers process trillions of bytes of data in 1000 hours   Said that the whole project took 10 years At present, Amunts and her colleagues have begun to devote themselves to the next human brain project focusing on individual differences, and she expects the new project to move forward faster   Understand brain data processing function   Perhaps the biggest challenge for human brain research is to understand how the brain stores and processes data The electron microscope information obtained by Jeff Lichtman of Harvard University and Winfried Denk of Max Planck Institute of neurobiology in Munich, Germany, shows that brain tissue can process 2000 terabytes in a cubic millimeter   It is estimated that a complete mouse brain can generate 60 petabytes of information, while a human brain can process 200 megabytes of information Lichtman said the amount of data is comparable to digital information across the world today, including Facebook and all other large data repositories   And that's just the beginning Neuroscientists will eventually collect the anatomical information of human brain (each brain is unique), and analyze the neural activities behind it layer by layer Scientists need to store and systematically organize these diverse data patterns in order to further explore the mysteries of the brain   Human brain engineering in Europe aims to provide a human brain simulation environment so that researchers can interact with it in real time, which is another level of demand "One of the challenges we face is to develop a computer language to make more efficient use of supercomputers," said Jesus labarta mancho of the Supercomputing Center in Barcelona, Spain, who is a partner in human brain engineering   At present, some experiments need to stimulate different parts of different brain regions at the same time Supercomputers can not meet this demand Therefore, it is necessary to develop a new method that enables supercomputers to compress part of the information in order to concentrate on the problems that need to be solved   Even if supercomputers can package and process data, theorists still need to think about what to ask them Christian machens, a theoretical neuroscientist at the champalimaud Center for the exploration of the unknown in Lisbon, Portugal, said: "this is a question of whether to have chickens or eggs first Once we understand how the brain works, we know how to process the data "   The scale and volume of the task are the focus of theorists' controversy; like many, kording believes there is too much to do "Google's search problem, by contrast, is like a kid's game," he said The number of neurons in the brain is roughly the same as the number of Internet pages But the problem is that Internet pages are connected in a straight line, and each neuron is connected with thousands of neurons, not in a straight line "   But Partha Mitra, a biological mathematician at Cold Spring Harbor Laboratory in New York, thinks that the biggest challenge to fully understand how the brain works comes from sociology "Tracking the brain works differently than tracking the Higgs boson, where each Higgs boson moves with a single target," he said For example, each individual in the team will speak freely when setting team goals Once the goals are set, they will strictly discipline themselves and work together to achieve the group goals "   As expected by Newsome, setting these project goals has taken him almost all summer At present, he is participating in a series of expert seminars aimed at setting research objectives for the U.S brain program, and revising a research report on the brain program to be released in September He said the report could not ensure that all brain related problems were solved, but would put research on the right track, making it possible to answer these questions in the future   "We will eventually be able to understand the impact of brain neuron activity on human behavior, which is crucial for us," he said