The virtual brain

How do you unravel the secrets of our very own thinking engine, and thus find a way to better treat illnesses of the brain? This is what neurologist Professor Henry Markram has been working on for over ten years in the context of his Blue Brain Project. And he is convinced that only relying on experiments and theories will not get us there.

Text: editorial staff “ceo” Magazine | Images: Alain Herzog / EPFL; Blue Brain Project / EPFL. All rights reserve | Magazine: Life & Science – July 2017

The human brain is still one of the biggest mysteries of humankind. On its outside, it is not a particularly spectacular, grey-whitish, soft matter of an average of 1.5 kilogrammes in adults. However, what goes on inside is absolutely fascinating: over 100 billion nerve cells or neurons communicate on a constant basis, build networks and synaptic connections, only to change them again if another function is required. And all of this happens in just a few milliseconds. The brain is the organ that controls all body functions, produces thoughts and emotions, and is responsible for the human consciousness. “The brain is a dynamic network of networks, operating in and across many levels at the same time,” Henry Markram explains. He is a professor of neuroscience at the École Polytechnique Fédérale de Lausanne (EPFL), director of the Laboratory of Neural Microcircuitry (LNMC) and the Founder and Director of the Blue Brain Project.

“I want us to understand the brain – preferably in my lifetime, not in that of my grandchildren.”

Founded in 2005 within the Brain Mind Institute of the EPFL, the aim of the Blue Brain Project is to build accurate, biologically detailed, digital reconstructions and simulations of the rodent brain and, ultimately, the human brain. In order to achieve this, the 55-year-old brain researcher follows a very unique approach. “This requires a new strategy, because by simply relying on experiments, we will never be able to understand the brain and how it works.” He is convinced that in the face of the incalculable number of reactions that take place in the brain, it will be impossible to conduct a sufficient amount of experiments. Added to this, the only time scientists ever get a look at a functioning brain is during an operation. This is why even after decades of research, mankind still knows very little about their own, complex thinking organ.

A billion-piece jigsaw puzzle

South African-born Markram is therefore all about simulations – a strategy that has already been used successfully in other scientific areas. Using a supercomputer, he and his team of scientists, software engineers and developers, technicians and postdoctoral researchers are developing the process of assimilating all the data and knowledge on the brain and constructing a computer model of the brain with which they can simulate the functions of the brain. In the first phase, they have been concentrating on a branched structure of thousands of neurons that measures but a few millimetres; a neocortical column of the cortex of a rat. This minimum unit of the animal’s brain serves as a basis for developing the entire organ and, in a later phase, an entire human brain.

For this purpose, the researchers rely on already published scientific data and facts on the brain and integrate it in the model. From this, they deduct rules and with them calculate the structure of the brain and the possible reactions of each individual cell. Markram draws a comparison with a jigsaw puzzle of a billion pieces, of which only about 1,000 pieces are known to us. “Normally one can’t build such a puzzle. However, by finding all the rules and interdependences between the pieces, we will be able to complete even such a gigantic puzzle.”

Henry Markram (55) is the Founder of the Brain Mind Institute, the Blue Brain Project and the EU’s Human Brain Project all head quartered at the EPFL. Born in South Africa, he studied medicine and neurophysiology in Cape Town. He immigrated to Israel and completed his PhD at the Weizmann Institute and his postdoctoral work at the National Institute for Health in the US and the Max Planck Institute for Medical Research in Heidelberg. The brain scientist is married and the father of five children. He and his wife are also conducting research on autism and founded Frontiers, which has become one of the largest Open Access publisher of science.

Failure produces insight

Using results from previous experiments, Markram and his team verify the deduced rules for their validity. For this, they consult approximately 100,000 scientific publications that are issued every year. If the assumed rules stand the test of all of these publications, the researchers see it as a confirmation that their assumptions are correct. It is, however, also interesting to encounter failure: “If something doesn’t work as expected, we know we are standing at the frontiers of our knowledge,” says Markram. “When one of our theories is disproved, we can question our assumptions in a systematic manner. We learn when we fail.”

“By relying solely on experiments, we will never be able to understand the brain.”

The Blue Brain Project published nearly 100 papers in the past ten years, culminating in the first draft digital reconstruction of neocortical microcircuitry published in the prestigious journal “Cell” (Markram et al, 2015). It is a detailed copy of a small piece of the rat neocortex, the most evolved part of the brain, that is approximately 1/3 cubic millimetres with 31,000 brain cells and 40 million connections between them. The researchers have since started building larger areas of the brain. Their next digital copy planned for the end of this year is the brain region responsible for the sense of touch and they are targeting the mouse brain with nearly 600 brain regions by 2023. In parallel that they have started on building a small piece of the human neocortex.

Knowledge is at the forefront

This was his vision, when Henry Markram brought the Blue Brain Project to life in 2005; “I want us to understand the brain – preferably in my lifetime, not in that of my grand-children,” he says. He is driven by the wish to find out how the brain influences a person’s own sensations and that towards their environment. He also has a personal incentive to understand what drives the brain, being the father of an autistic son. “Even as a neurologist, I feel powerless,” he says. Healing brain illnesses is still very much of a gamble today. “There is a great deal of trying out. If something works, it is going to produce a new enterprise worth billions, or, if it doesn’t, you go back to the drawing board.” The computer model of the brain is intended to provide further fundamental data for this and improve the experiments and trials. For example, the Blue Brain Project could help locate particularly vulnerable parts of the brain and identify strategically important areas. “Whoever treats illnesses of the brain absolutely has to know how it works,” he says. In the future, Markram hopes that the simulations with the computer model will help reduce the 600 different illnesses of the brain known to us to date.

The aim of the Blue Brain Project, a Swiss brain initiative founded and directed by Professor Henry Markram, is to build accurate, biologically detailed digital reconstructions and simulations of the rodent brain, and ultimately, the human brain. The supercomputer-based reconstructions and simulations built by Blue Brain offer a radically new approach for understanding the multilevel structure and function of the brain.

An advantage for artificial intelligence

The calculations for the model are carried out by supercomputer Blue Brain IV IBM BG/Q. It will soon be replaced by the next generation of supercomputers with even more processing power. The idea that these supercomputers will one day start to think by themselves, however, is purely science fiction for Markram – at least for the next 100 years. However, simplified knowledge gained in the project could find a way into our everyday lives earlier and have a significant influence on the research on artificial intelligence. “The brain works much more efficient, faster and less data-intense than today’s computer networks,” says Markram. In the next few years, a number of spin-offs in the area of artificial intelligence are expected to emerge from the project.

The fact that this ambitious project is based in Switzerland is, according to Markram, the result of the favourable conditions he is being offered. A particularly important prerequisite in order to launch this long-term research in Switzerland was the ETH Board’s and the Swiss Federal Council’s backing of the Blue Brain Project and its vision. “Our team is very interdisciplinary. Our members are often required to acquire a considerable amount of knowledge in a field that is partially alien to them in order to understand the entire model,” Markram explains. He is determined that Switzerland should be rewarded for its commitment. “Thanks to its leading role in this project, Switzerland will be in a key position to develop new treatments for the brain,” Markram continues. He also expects positive consequences for information technology and communication technologies.

And it may not be that long until this happens, at least according to the brain researcher. The problems that are still in the way of this are clearly defined and identified. “By the end of our funding period in 2023, we will certainly be able to present a detailed simulation of the brain of a mouse and possibly a lower resolution model of the human brain,” he says. And no doubt on the way there, we will discover a great deal of other amazing facts about the brain.

Image: Blue Brain Project / EPFL © 2005 – 2017. All rights reserved

The complexity of the neocortex. The complexity of the brain is portrayed here showing three major components that interact with each other in the neocortex (the most evolved part of a mammal’s brain): blood vessels, astrocytes (supporting cells) and neurons.
Image: Blue Brain Project / EPFL © 2005 – 2017. All rights reserved

The neocortical microcircuitry – a view from inside the neocortex of the highly organised network of neurons.