Humphrey Obuobi ’18, THURJ Staff

“Our mission is to build a detailed, realistic computer model of the human brain.”

With this statement, any remaining murmurs in the crowd faded to a hush. For the next fifteen minutes, Henry Markram, a well-recognized neurobiologist at the time, hooked his audience with aspirations of optimizing relationships and curing the world of mental diseases through high-throughput simulations (7). The end result he envisioned — a hyper self-conscious human race with the ability to predict the slightest nuances and changes in the brain — seems far-fetched, but Markram assures us that the time to take control of our own minds is now. His words did not pass by idle ears, either; with the popularity of this talk came the conception of Markram’s “brain child” — the Human Brain Project, a 10-year, 1.2 billion euro effort with the sole purpose of achieving the goals that Markram proposed (10). The effort has been hailed by many as the next Human Genome Project, with the question of brain function being one of the hot topics of scientific research this decade.

Markram isn’t the only scientist in the field to make such a bold claim about the brain. With our knowledge advancing alongside the speed of technology and our computing capabilities, many bold individuals have looked into the integration of mind and machine. Randal Koene, a Dutch computational neuroscientist, made a splash at Transhuman Visions 2014 when he suggested that the mind — that being the brain and all its internal connections — could be uploaded to a computer (11). In a similar fashion to Markram, he emphasized the computational nature of the brain itself, reducing the action potentials and synapses to complex, yet concrete physical equations. In this way, Koene believed that his mind could survive indefinitely.

But what is it that makes these two men in particular so confident in the intimate connection between man and machine?

The Course of Computational Neuroscience

Once the academic community accepted Santiago Ramon y Cajal’s neuron doctrine – the framework that identifies neurons as the functional units of the brain – as the foundation of modern neuroscience in the late 1800s, our understanding of the brain rocketed forward. In 1939, Kenneth S. Cole and Howard J. Curtis of Columbia University published a paper titled “Electric Impedance of the Squid Giant Axon During Activity” that analyzed the changes in transmembrane voltage of a giant squid axon, thus determining the general character of what we now know as the “action potential” (Cole et al. 1939). The paper laid out the foundation for the neuron (and the rest of the nervous system) to be modeled according to physical principles. Shortly after this paper was published, Hodgkin and Huxley used their separate understandings of biology and electronics to develop the neurophysiological equations known (quite appropriately) as the Hodgkin-Huxley model, taking into account ion channel permeability, capacitance of the lipid bilayer, and many other factors (9). In this way, the community of neuroscientists moved beyond a vague hypothesis to an actual mathematical model, giving researchers the power to predict.
Over the past sixty years, the field has advanced at a breakneck speed; where the centuries prior had focused primarily on coming to a consensus on the nature of the nervous system and its ailments, the latter half of the 20th century began to investigate the nervous system with an unprecedented level of detail. Techniques such as functional magnetic resonance imaging (fMRI) allowed scientists and physicians alike to trace the pathways of activity in the brain, while voltage patch clamp experiments more clearly illuminated the electrical behavior of the individual neurons. The greatest shift in thought was not technological, however; rather, one of the most prominent advances in neuroscience in the last fifty years has been a shift in focus to connectivity. Thus, pathways began to dominate individual neurons as the probable explanations for all animal and human behaviors. Technologies such as diffusion tensor imaging (DTI) have now given scientists the capability to image these connections in visually stunning ways. Relating the connections of the brain to its mechanistic function, our idea of the brain slowly became that of an incredibly complex machine, but a machine nonetheless.
For the average computational neuroscientist, this new focus on connectivity and the parallel advances in technology meant more accurate predictions of the brain’s function and “smarter” artificial intelligence; the projects from the field have ranged from natural language processing to smarter search engines. However, for figures like Koene or Markram, this increase in understanding indicated that the mind and all of its facilities could be modeled to a new and fantastic degree of precision. Everything from deciding what to eat for breakfast to solving complex integrals might be just a few computations away. Now, the focus needed to be set on actually going through with this idea.

Mission for the Mind

The Human Brain Project (or HBP), established in 2013, was essentially created to do just that — to use computer simulations to mimic the physiological conditions and computations of the brain. Founded by none other than Henry Markram, the project aims to “use neuroinformatics and brain simulation to collect and integrate experimental data, identifying and filling gaps in our knowledge, and prioritising future experiments” over the next ten years (7). The project has had no dearth of money; currently, it is being funded by the European Union and several other donors, bringing the invested money up to a grand total of 1.2 billion euros.

Overall, the HBP is focused on integrating the brain and the computational power of supercomputing to facilitate astronomical advances in neuroscience, medicine, and computing. While not all of the methods to be implemented in the project are available to the public, the official public release of the goals of the HBP provides some insight into the methodology of the research. Reminiscent of classical neurobiological research, the project plans to use mouse brains as a necessary stepping stone to understand the human brain, relying on genomic sequencing, optogenetics, molecular systems biology, and a slew of other research techniques to map the connections of the mouse brain and later extrapolate the results to the multi-level structure of the human brain (8). Predictive neuroscience was also stated to be used to “fill the gaps in the data,” as not all the parameters could be determined.

In the first year of what the HBP calls its “ramp up stage”, there has been quite a bit of fundamental progress. At the Year One Summit for the HBP, achievements were divided according to the level of organization; for instance, SP1 denoted achievements at the cellular and molecular level, while SP2 covered advances on the frontier of brain structure and function. And for one year of research, the results aren’t bad at all. Researchers at the École Polytechnique Fédérale de Lausanne and other involved institutions have already determined figures for synapse density (crucial for determining cellular connections in the brain), illuminated circuits relating to self-awareness, and much more (5). The project is still young, though, and these advances must be taken with a grain of salt; to realize the same degree of progress in the upcoming years will take sustained effort by all parties involved.

Too Much, Too Early

Markram’s enthusiasm for the future of brain science may be contagious, but it is quite possible that his grandiose visions of neural computing and brain simulations may be too far ahead of his time.

The project has consistently been under fire from the majority of the neuroscience research community, who have criticized the project and Professor Markram for “siphoning funds from fundamental research” and have pointed out that “large-scale simulations make little sense unless constrained by data” (4). The concerns are wholly legitimate; intuitively, it makes sense that in order to model a system, one has to have enough parameters and details of the system, grounded in repeated experimentation. The outrage of the neuroscience community has been further exacerbated by the Human Brain Project’s decision to reduce the same fundamental experimentation that will clarify these very parameters. Why? One can only imagine that the goal of such a move would be to save time and money in their endeavor to map the human brain.

Given all of these obvious hints at inefficiencies in the structure of the Project itself, a community of neuroscience researchers, a group of 750 supporters both in Europe and abroad, have written an open letter to the European Commission, pointing out the flaws of the current design. The letter addresses the “transparency and accountability and… governance” as points to be improved in the Human Brain Project, stating clearly: “We believe the HBP is not a well conceived or implemented project and that it is ill-suited to be the centerpiece of European neuroscience.”
The European Commission was quick to respond. In their response letter, the Commission agrees with many of the points raised by the attacking party, encouraging written suggestions for the improved governance and structure of the program (6). In addition, future collaboration with the United States’ BRAIN Initiative was mentioned as a potential path to improve the system; this Initiative’s focus on experimental studies and simply mapping the activity (rather than connectivity) of each neuron in the brain could perhaps provide the necessary foundational data to create a feasible model (1). However, nothing concrete has been done as of yet to move in that direction. As the Commission holds, there is “no single roadmap for understanding the human brain.”

* * * * *

Despite the inherent problems with its system, the Human Brain Project and efforts like it remain exciting endeavors on the cutting edge of neuroscience — the border between man and machine. If anything, these projects should make it clear how far the scientific community has come from even the beginning of the century; as technology has barreled forward into the future, our knowledge of the brain and its functions have also leapt to great heights. Perhaps the time isn’t now to say that the two paths have become indistinguishable, but projects like the Human Brain Project make it clear that the scientific community is ready to invest in a computational approach to understand our minds.
And who knows? Maybe scientists will be able to model your brain sometime in your lifetime.

References

  1. “Brain Research through Advancing Innovative Neurotechnologies (BRAIN).” Brain Research through Advancing Innovative Neurotechnologies (BRAIN). National Institutes of Health. Web. 16 Nov. 2014. <http://www.braininitiative.nih.gov/index.htm>.
  2. Caramenico, Greg. “EconoMonitor : EconoMonitor » Irrational Funding Exuberance: Why the ‘Brain Initiative’ Is Not the Next Human Genome Project.” EconoMonitor. EconoMonitor, 23 Aug. 2013. Web. 16 Nov. 2014. <http://www.economonitor.com/blog/2013/08/irrational-funding-exuberance-why-the-brain-initiative-is-not-the-next-human-genome-project/>.
  3. Enserink, Martin, and Kai Kupferschmidt. “Updated: European Neuroscientists Revolt against the E.U.’s Human Brain Project.” Science/AAAS. AAAS, 11 July 2014. Web. 16 Nov. 2014. <http://news.sciencemag.org/brain-behavior/2014/07/updated-european-neuroscientists-revolt-against-e-u-s-human-brain-project>.
  4. Frégnac, Yves, and Gilles Laurent. “Neuroscience: Where Is the Brain in the Human Brain Project?” Nature. Nature Publishing Group, 3 Sept. 2014. Web. 16 Nov. 2014. <http://www.nature.com/news/neuroscience-where-is-the-brain-in-the-human-brain-project-1.15803>.
  5. “HBP Achievements: Year One.” Human Brain Project. European Commission, 29 Sept. 2014. Web. 16 Nov. 2014. <https://www.humanbrainproject.eu/documents/10180/538356/HBP Achievements Year 1/d0dc08df-cdc7-4242-a677-eef632f622c3>.
  6. Madelin, Robert. “No Single Roadmap for Understanding the Human Brain.” Digital Agenda for Europe. European Commission, 18 July 2014. Web. 16 Nov. 2014. <https://ec.europa.eu/digital-agenda/en/blog/no-single-roadmap-understanding-human-brain>.
  7. Markram, Henry. “A Brain in a Supercomputer.” Henry Markram:. TED, 1 July 2009. Web. 16 Nov. 2014. <http://www.ted.com/talks/henry_markram_supercomputing_the_brain_s_secrets?language=en>.
  8. Markram, Henry, and Karlheinz Meier. “Science and Technology Plan.” The Human Brain Project: A Report to the European Commission. Lausanne: European Commission, 2012. 28-57. Print.
  9. Moore, J.W., and M.L Hines. “Chapter 2. A Brief History of Computational Neuroscience – – From My Perspective.” Chapter 2. A Brief History of Computational Neuroscience. Duke University, 19 Oct. 1995. Web. 16 Nov. 2014. <http://neuron.duke.edu/userman/2/pioneer.html>.
  10. “Overview.” Human Brain Project. European Commission. Web. 16 Nov. 2014. <https://www.humanbrainproject.eu/vision>.
  11. Piore, Adam. “The Neuroscientist Who Wants To Upload Humanity To A Computer.” Popular Science. Popular Science, 16 May 2014. Web. 16 Nov. 2014. <http://www.popsci.com/article/science/neuroscientist-who-wants-upload-humanity-computer>.

Comments:

NO COMMENTS

LEAVE A REPLY