Blue Brain: modeling the brain
Humanity was promised an artificial brain
Alexey Timoshenko, GZT.RUAn artificial brain will be created within the next 10 years.
This was stated at the TED Talks conference by Henry Markram, the head of the Blue Brain project.
Within the framework of the Blue Brain project, which began in 2005, it is supposed to recreate one of the elements of the cerebral cortex using computer modeling methods. In mammals, including humans, it is the cerebral cortex that is responsible for such important functions as thinking, speech and processing information about the outside world.
Processors instead of neurons
Scientists from the Swiss Scientific Center have already modeled 10 thousand artificial neurons and connected them into a single network: for this purpose they had to use the Blue Gene supercomputer. Despite the fact that in reality the brain is disproportionately larger (100 billion nerve cells, that is, 10 million times more than scientists have already created), 10 thousand neurons are already quite a lot, and for some studies it is quite enough.
The Blue Brain project is not just about a random network of nerve cells. Henry Markram's team is building a model of a neuronal column. There are such columns in all parts of the cortex, and at the same time they are quite versatile: in the visual cortex they can process an image, in the motor cortex they can control the movement of hands, and in the temporal cortex they can form associative connections.
If researchers manage to get a working model of one of the basic blocks of the mammalian brain, it will help not only neuroscientists, but also doctors. By changing the parameters of the model, it will be possible to reproduce the symptoms of the disease and track the effect of certain drugs. It is now known how certain medications affect the connections between nerve cells. But how this change is reflected in a separate part of the brain and why it is like this at all – these questions remain unanswered.
Trillions of combinations
The main problem that Markram says his team has encountered is determining the connections between neurons. A real nerve cell has thousands of contacts, which in a column of 10 thousand cells gives an astronomical number of possible options.
In addition, it should be added that the same neuron, due to different neurotransmitters that provide the signal transmission of substances, can use several types of connections independently of each other. Nerve impulses can also lead to both activation and inhibition of a neuron. In addition, the neurons themselves come in several different types. All these many factors explain the reason why the work started in 2005 can be completed no earlier than in 10 years.
It should be noted that Markram himself began to create an artificial brain long before the project. In his report, he noted that he has been describing real neurons for 15 years: "It's like the work of a botanist in a tropical forest, who describes various types of trees with their branching crowns. But our work is complicated by the fact that we are still studying the connections between neurons and trying to understand the rules by which these connections are built."
From molecules to the brain
In November 2007, the first stage of the Blue Brain project was completed. Then scientists managed to build the first, still very imperfect model of the neuronal column. Nevertheless, this made it possible to continue working in two directions – down to the molecular level, and up to the brain as a whole.
The "down" level implies taking into account molecular processes inside nerve cells. Each cell in the model requires a separate processor for modeling: if once, at the dawn of neuroscience, scientists considered a neuron as a simple node of a nervous network, then modern experimental data describe it as almost an independent organism.
This image shows real neurons in the mouse cerebral cortex.
The numbers on the edges are a scale in micrometers, a pyramid neuron with many connections is shown in close–up.
Pay attention to the frame around the picture: modern microscopes allow you to get a three-dimensional image.
Sitting in front of the microscope monitor, the researcher can rotate the object for viewing from different sides!
www.advanced-imaging-center.org, Miguel Vaz Afonso
In response to nerve impulses, neurons begin to change the work of their genes: various proteins are synthesized in them, connections with neighboring cells change – all this can radically change the work of the brain as a whole. To illustrate the importance of such processes, let's say that sometimes their result is even the death of a neuron or the birth of new cells: the opinion that nerve cells do not recover has also been refuted in recent decades. Within the framework of the Blue Brain project, it is supposed to analyze how the inner life of a neuron, individual genes and molecules affect the work of neuronal columns.
On the other hand, when moving up, the presence of a separate neuronal column will also help in modeling the whole brain. It is unlikely to create as many columns as there are in the brain of at least a mouse, but even more primitive models can significantly advance medical research: they can be used to read complete information about any "cell", about all its connections, about the concentration of certain substances.
If it is difficult to achieve accuracy in research in experiments on a live mouse (you have to kill an animal, and then make thin slices of the brain, conduct complex and expensive analyses), then these experiments are impossible in principle on a human. In the event that a group of scientists involved in the Blue Brain project succeeds in eventually bringing the model to the desired level, these studies will be able to help two billion people. That's how many, according to Henry Markram, are now suffering from various brain diseases, Alzheimer's disease, schizophrenia and similar ailments.
Portal "Eternal youth" http://vechnayamolodost.ru24.07.2009