Shulou(Shulou.com)11/24 Report--

Ten years, the European Union's human brain project is about to expire, but after burning up 4.7 billion yuan, the results are still fragmented, and the grand plan to rebuild the human brain still seems out of reach.

Ten years ago, European scientists planned to reconstruct the human brain in supercomputers. Ten years later, most of the EU's 600 million euros have been wasted.

The European Union has invested nearly 4.7 billion yuan in this plan called Human Brain Project (HBP for short). There are more than 500 top scientists who have published thousands of papers.

The 10-year human brain project will end in September this year. However, for human beings, a full understanding of the human brain is still out of reach. Not to mention the use of computers to simulate the human brain and develop brain-like intelligence.

Just today, Nature sent a long article detailing the achievements of the project and the complex and turbulent process.

Article address: https://www.nature.com/articles/d41586-023-02600-x

Swiss scientist: simulating the human brain is not impossible. The earliest sponsor of this project is Henry Markram, a Swiss neuroscientist who boasted about it more than a decade ago.

In 2009, at the British (TED) Congress, he announced a plan that shocked everyone:

On the basis of understanding the structure of the brain, he wants to use a computer to create a complex mathematical model that simulates 86 billion neurons and 100 trillion synapses in the human brain.

Perhaps out of excessive optimism about computer technology, he thinks it is not impossible to simulate the human brain.

If we can successfully simulate the human brain, it will be of great significance to humans: Alzheimer's disease can be overcome, robots will be smarter, more cognitive, and even speed up computers.

As a result, the project received a lot of attention and attracted a huge amount of money after it was announced. It can be said that the European Union has spent a lot of money on HBP.

In 2009, for fear of lagging behind the United States in technologies such as computers and digital services, the European Commission's General Administration of Communications Network content and Technology set up a "Future emerging Technology flagship Project", which will receive 1 billion euros in funding.

In 2013, the "Human brain Project", which is expected to be 10 years old, was officially launched, led by the European Union government and with the participation of 135 cooperative agencies from 26 countries.

The project, with a planned investment of 1.3 billion euros, aims to develop an information and communication technology platform dedicated to neuroinformatics, brain simulation, high-performance computing, medical informatics, neuromorphological computing and neurorobotics.

The project will focus on simulating brain function through supercomputer technology to achieve super artificial intelligence.

Some achievements have been made in the digital reconstruction of the neuronal circuits in the temporal lobe of the human brain, and there has been a lot of criticism that the HBP project has made some achievements within a decade. With the publication of thousands of papers, great progress has been made in the field of neuroscience.

Scientists have created detailed three-dimensional maps of at least 200 brain regions, developed brain implants to treat blindness, used supercomputers to model functions such as memory and consciousness, and promoted the treatment of various brain diseases.

However, the human brain project has not been favored since its inception and has been criticized.

Thomas Skordas, director of the European Commission, said: "at the beginning of the project, few people believed in big data's potential, let alone that humans could use supercomputers to simulate complex brain functions. "

In the eyes of many scientists, this goal itself is unrealistic. Not only that, HBP has repeatedly changed the direction of research, making scientific research output fragmented, like mosaics.

Yves Fr é gnac, a cognitive scientist at the French National Research Center (CNRS), said HBP failed to provide a comprehensive or unique understanding of the brain. "I can't see the whole brain, I can only see some fragments of the brain," he said. "

Perhaps the biggest achievement of the project so far is a digital virtual platform called EBRAINS, which allows scientists around the world to do simulation experiments.

"Today, we have all the tools to build a digital brain twin," said Viktor Jirsa, a neuroscientist at the University of Aix-Marseilles in France and a member of the human brain project board. "

The heads of the Human brain Project hope that the platform will move further towards their ambitious goals. But the source of funding for the project is still unstable. Against the backdrop of such a huge investment, the slow progress of the project has frustrated European scientists.

"We may be the first to charge the brain," said Jorge Mejias, a computational neuroscientist at the University of Amsterdam who joined HBP in 2019. But now that every team in the world is racing against the clock, we don't have time to stop and rest. "

The tortuous course of the human brain project has been controversial from the beginning.

One of the main goals of the project when it landed in 2013 was to develop the tools and infrastructure needed to study the brain to better understand its function, organization, and the diseases associated with it. It also includes small projects on basic and clinical neuroscience.

At that time, another key project alongside it was the study of graphene.

The European Union and project financiers have pledged 1 billion euros in financial support for the human brain project. In the end, it received 607 million euros in research funding, of which 406 million euros came from the European Union.

These funds are released in four stages and are gradually provided to laboratories that compete for funding at each stage.

But in the first year, the project ran into trouble.

Henry Markram, the neuroscientist who proposed the project in TED's speech, declared publicly that HBP will be able to reconstruct and simulate the human brain at the cellular level within a decade.

His assertion aroused widespread suspicion among neuroscientists. But Markram responded: "when science plans a new path, controversy will naturally follow." "

Timothy O'Leary, a computational neuroscientist at the University of Cambridge in the UK, believes that this ambitious goal will help kick-start HBP. Because, "if you don't come up with an ambitious or even ridiculous goal, the human brain project may not necessarily be funded." "

With the passage of time, the leadership of Markram has been more and more questioned.

In 2014, he and two other members of the executive committee changed the focus of the project, cutting off a large number of research in cognitive neuroscience, causing 18 laboratories to withdraw from the project. However, Markram believes that there is a dispute among various departments on the allocation of funds.

More than 150 scientists then signed a letter of protest urging the European Commission to reconsider the purpose of HBP in a timely manner and to conduct a second round of funding. The letter expounds the current situation that HBP is poorly managed and has partially deviated from the scientific track.

Instead of responding directly to the problem of mismanagement, Markram pointed the finger at some scientists who disobeyed him:

It is clear that some people in the neuroscience community are not ready to unite under a unified vision. "

As a result, the EU established a committee of independent experts to study how the project works and to revise its objectives.

The Committee recommended that HBP reassess and articulate its scientific objectives more clearly and reintegrate cognitive science and systems neuroscience into its core programmes.

In February 2015, HBP's board voted to dissolve the three-member executive committee and set up a larger committee to replace it.

The upheaval has caused resistance among many scientists to the project. "this skepticism has been interfering with the development of the project," Mejias said. "

At the same time, large-scale brain projects around the world have also started or entered the stage of high-speed operation. Both the United States and Japan launched brain research projects at the same time as HBP-the United States will continue until 2026 and Japan will continue research for 15 years.

China's brain research project was launched in 2021, and projects in Australia and South Korea have entered their seventh year.

But the capture of the project did not end with the dissolution of the executive committee.

From 2016 to 2020, there have been many changes in project management. At this time, the project research finally began to speed up.

In 2016, as a result of the development phase of the project, HBP launched six professional operating platforms covering areas such as brain simulation, high-performance analysis and computing, and neural robotics.

"our idea is to integrate these six directions over time, but at the beginning, they are quite independent," said Katrin Amunts, a neuroscientist at the J ü lich Research Center in Germany and director of HBP. "

"A big project like HBP is often accompanied by a learning process, and when the project was first started, a lot of things were not on track. "

Although it is a 10-year period of 600 million euros, but after this period of trouble, the effect of the money and time invested in the early stage has been greatly reduced.

Important achievements aside from management, HBP has accumulated a lot of key scientific research achievements.

By creating and combining about 200 3D maps of the cerebral cortex and deeper brain structures, HBP scientists created a map of the human brain-depicting multiple layers of brain tissue, from cellular and molecular structures to functional modules and connectivity.

At the 2023 HBP Summit in March, Amunts said: "the map of the human brain is a bit like Google Maps, but the Google Maps of the brain. "

The map uses postmortem brain data to generate standardized maps and takes into account natural differences between people.

Using this map, HBP scientists identified six previously unknown areas of the brain in the prefrontal cortex that contribute to memory, language, attention and music processing.

At the same time, after linking the map to the gene expression data in Allen's Human brain Map (Allen Human Brain Atlas), it can be used to characterize neurons in the entire brain.

Using these paired maps, the researchers revealed how changes in gene expression associated with depression are associated with structural and functional changes in the frontal cortex.

In addition, HBP researchers have developed a unique algorithm that can build full-size scaffold models of brain regions based on microscopic images.

Using this tool, the researchers created a detailed map of the CA1 area of the hippocampus, which is important for memory. It contains about 5 million neurons and 40 billion synapses.

At present, HBP has translated some of the findings in the study into clinical application. For example, personalized brain models, known as "digital twins", can be used to improve the treatment of epilepsy and Parkinson's disease.

Not all, however, HBP has a long-standing problem in neuroscience research: fragmentation. This is also a common criticism of the project's organizers and critics.

As Fr é gnac said:

"you can see very ingenious designs and inventions, but you can hardly see multi-scale integration, nor can you see the solution to big problems. "

HBP has described in detail the anatomy of the human brain and developed tools to link brain structure and function to gene expression, so over the past three years, HBP has been trying to overcome the fragmentation of its interdisciplinary subprojects.

Project participants have also made a series of efforts to integrate technologies into EBRAINS and HBP's six platforms advocate the development of compatible tools and shared data standards. Some groups have also been restructured and future work will focus on specific scientific challenges rather than disciplines.

But Jirsa said: "there is still a lot of work to be done." Neurorobots still have no connection to clinical-oriented groups (clinically driven group). "

Some researchers believe that the "lack of focus" has led to the fragmentation of HBP's scientific achievements.

Fred Wolf, a theoretical neurophysicist at the University of Gottingen in Germany, left HBP after signing the open letter. "I hope this project, which has been going on for more than a decade, will produce a conceptual breakthrough," he said, but this is not the case with HBP.

David Hansel, a neuroscientist at the Paris Center for Comprehensive Neuroscience and Cognition, who is not involved in the HBP project, believes that HBP lacks priorities and cooperation is limited.

This means that HBP has failed to take advantage of its own size and has failed to really unite the neuroscience community under a common goal.

"the project does not list the top and reasonable issues to be addressed," Hansel said. Simply set a goal of "understanding the brain".

But John Ngai, director of the brain Research by promoting innovative Neural Technologies (BRAIN) Program at the National Institutes of Health in Bethesda, Maryland, believes that science that emphasizes data collection rather than hypothesis-driven is tenable.

Ngai said, "Big Science is not always about landing on the moon, especially when the steps to achieve the main goals are uncertain." "

At the end of September this year, HBP will completely stop issuing project funds.

Although some of the results from HBP have been funded to continue their work, the future remains bleak for most or all of the researchers who work with HBP.

The HBP project went bankrupt, but Amunts and others hope that HBP's work and the EBRAINS platform will be the foundation of European neuroscience in the coming years.

"the study of the brain requires understanding the multi-levels and multi-scales of the brain," Amunts said. "

In January 2018, HBP received funding of 50 million euros, including 25 million euros from the European Union, for the development of interactive supercomputing tools and data storage services for EBRAINS. Researchers have begun to use the platform to study how the brain responds to stimuli and to develop brain-like robots.

Ngai says HBP's move to EBRAINS has given birth to a valuable tool. Although there are similar platforms, they lack the scale and services provided by EBRAIN.

In March, the European Commission rejected an application for 38 million euros to keep EBRAINS running. But after consulting with HBP, the committee resumed the same fund-raising in June, giving the team another chance to apply.

If this application also fails, then HBP will have to rely on private funds and financial support from EU countries to maintain its operation.

At the same time, the European Commission is preparing to evaluate the HBP project. It is reported that the final review of HBP will begin in November, and the results will be announced in January 2024.

"if we don't want to live like the winter of artificial intelligence in the global neuroscience field, we need to make this field respected," Fr é gnac said. At the same time, we really need to assess whether such flagship projects are good or not. "

EBRAINS's CEO and HBP Director-General Pawe HBP wieboda also said that the end of HBP is not the end of European neuroscience.

The European Commission and member states are planning the next phase of the European brain Health study, focusing on the use of personalized brain models to advance drug development and improve the treatment of brain diseases.

But researchers say future projects need to avoid the pitfalls that HBP has stepped on.

"We don't want to repeat another HBP project like we did," O'Leary said. "We need to support small-scale, focused scientific research and ambitious integrated projects. "

It is undeniable that a group of scientists who focus on common goals have been born in the HBP project. And this will be an eternal treasure.

Reference:

Https://www.nature.com/articles/d41586-023-02600-x#ref-CR3

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.