Nature sub-journal transcends Nobel classic theory: Neuroscience research is a long way off...

|Original compilation, reprint must indicate the source! Scientists are looking at a two-photon microscope used to record the activity of brain cells in mice.Image Source: a new study on the activity of nearly 60000 neurons in the visual system of mice published in nature neuroscience on December 17, Beijing time by Allen Institute shows that we still have a long way to go to understand how the brain calculates and how to process sensory information to guide behavior.the analysis led by the Allen Institute of brain science in the United States shows that more than 90% of neurons in the visual cortex, which processes the visual world, do not work as scientists think, and how they work is still a mystery.Dr. Christof Koch, chief scientist and director of the Allen Institute of brain sciences, said: "we believe that the principles of these neurons in processing visual information are simple, and they are found in all textbooks.but now we've examined thousands of cells at the same time, and we've got a much more subtle and complex image."he and Dr. R. clay Reid, a senior researcher at the Allen Institute of brain sciences, are co authors of the study.in 1959, two neuroscientists, David Hubel and Torsten wise, conducted an experiment on how the mammalian brain perceives the surrounding visual world, which became a breakthrough and a real turning point in the field of neuroscience.their study revealed a single neuron that responded only to a specific type of image.two scientists, David Hubel and Torsten Wiesel, did their feat by showing simple pictures to cats and monkeys, such as black bars or dots on a white background.the basic principle they reveal is that when you look at the world around you, specific neurons in the brain are responsible for identifying exact parts of a specific area in the scene, and this recognition becomes more professional and sophisticated in the higher-order parts of the brain.suppose you're in the park: a group of neurons respond quickly to a dark branch in your line of sight.other neurons are activated only when the bird flies over your field of vision from left or right.then, your brain stitches together the information from the "branch" neurons and the "bird" neurons to get a complete picture of the world around you, at least in theory.the discoveries of Hubel and Wisel have been recognized by the Nobel Prize in physiology or medicine, and have formed the basis of neural networks supporting most applications of computational vision.in the past decade, with the advent of new neuroscience methods, more and more brain cells can be studied at the same time.scientists have come to understand that this model of our brain does not seem to be complete, that is, some neurons obviously do not follow the classic model of regulating to specific functions.but it is not clear how incomplete the model is.variability in brain activity this new study is the first large-scale analysis of open data from the Allen Brain Observatory (ABO), a large-scale physiological survey of the visual cortex of 243 awake mice, capturing the activity of tens of thousands of neurons in their visual system.the researchers analyzed the activity of nearly 60000 different neurons in the visual part of the outermost cortex of the brain when animals saw different simple images, photos and short video clips.the video section includes the opening shots from the classic Orson Welle movie touch of evil.it was chosen because it has continuous motion and is a single shot without cutting.source: Nature Neuroscience, neuroscience in the 1950s and 1960s was like fishing expeditions.at that time, researchers searched the brain with an electrode until they found neurons that responded reliably to specific images.Koch said: "it's like trying to see a wide screen movie through a few scattered pinholes, and it's impossible to get a full picture.the Allen Brain observatory data set does not capture the activity of each neuron in each case, but it allows researchers to study more neurons at a time, including those with more subtle responses."the researchers' new analysis found that less than 10% of the 60000 neurons responded to the textbook model.in the rest, about two-thirds showed some reliable responses, but obviously exceeded the predictions of classical models.the last third of the neurons showed some activity, but they did not respond stably to any stimulus in the experiment, and the role of these neurons was not clear.Dr. Saskia de Vries, co-author of the study, said: "this is not to say that previous studies have been completely wrong, but that these cells only account for a small proportion of all neurons in the cerebral cortex.our results show that the visual cortex of mice is much more complex and rich than previously thought, which also emphasizes the value of doing this research. The visual activity of the mice was recorded by two-photon microscope.Image Source: the presence of these variable and low specificity neurons at the Allen Institute is nothing new.but surprisingly, they dominate the visual part of the mouse brain, the researchers said.how the brain calculates is unclear how these other neurons process visual information. previous research teams have found that exercise can drive the activity of neurons in the visual part of the brain, but whether mice are running or not, it can only explain a small amount of changes in visual response. their next research will use more natural documentaries to carry out similar experiments, so as to provide neurons with more visual features to respond. the researchers also pointed out that the classic model comes from the study of cats and primates, both of which have evolved to see their own world better than mice. it is possible that the visual system of mice is completely different from ours, but there are still some principles in these studies that may apply to our own brains. Michael buice, co-author of the study, said: "our goal is not to study vision, but to study how the cerebral cortex is calculated. we believe that the cerebral cortex has a general computing structure, similar to the way that different types of computers can run the same program. at the end of the day, it doesn't matter what kind of program the computer is running. We want to know how the program works. ”