The Blue Brain Project- An Overview

What it is and the technological progress so far

The Blue Brain project is basically a research initiative to create an artificial digital rendition of a human brain. It sounds impossible, like a sci-fi movie, doesn’t it?

Before we get to describing how exactly it is being done, let’s do something fun, shall we?

Take your right hand and pinch your left forearm. Felt the pain?

Ever wondered why do we feel that?

Let me tell you (and even those who were too mature to do it). The whole process consisted of lifting your right hand, then taking the index finger and thumb, pulling a bit skin of your left forearm into the cavity formed, and pinching it. All of this (including the pain felt) was controlled by a small, just 2 mm patch of 6-layer brain tissue, called the Neocortex.

But how were the impulses generated and sent to the brain to make you feel pain?

This is what interested many of the neuroscientists and due to advancements in the field, we now know why it happens. All credits to the network of neurons residing in our brain. But can we show the whole thing on a computer? Virtually? With all the neuron activations as well as the tiny sparks transferred from one to another.

A guy, in 2005, told the Brain and Mind Institute of École Polytechnique Fédérale de Lausanne (EPFL) in Switzerland that, “Yes, we can” and said he can lead the project with proper funding. He was Henry Markram and went on to lead the team of numerous scientists for over a decade. The project was named the Blue Brain Project. The project is funded at the rate of 20 million dollars per year by the European Council. Along with Henry, there were two more people, who were going to co-direct this project. Felix Schürmann and Sean Hill. It started with the purpose of mapping the fully functional brain of a rodent on a computer, which will then be scaled up to a fully functional mapping of the human brain.

A computer simulation of the upper layer of a rat brain neocortical column. Here neurons light up in a “global excitatory state” of blues and yellows. Courtesy of Dr. Pablo de Heras Ciechomski/Visualbiotech

We all know that the brain has numerous amounts of connections and neurons to deal with and it is not possible for a commercial computer to compute that much. Hence, IBM came into the picture. Blue Gene supercomputers were the first ones to take this journey ahead with their petaFLOPS computational speeds. The supercomputers will be running Michael Hines’s NEURON software, which is used to model simulations of a network of neurons.

Before we discuss the technologies and how they developed these, let’s see how much progress they have achieved.

• 2006, the team generated its first model of a cortical column.
• 2007, they completed building the full cortical column of a rat.
• 2010, they scaled up the project from the rat brain to the human brain. The project was given an obvious name “The Human Brain Project”.
• 2012, the team developed methods that will make it possible to predict the connectivity of neocortical microcircuitry.
• 2013, the team developed models to identify and characterize neuronal clusters in neuron microcircuits.
• 2015, a study was released which proved the feasibility of building and stimulating a digital copy of the brain, which was a breakthrough.
• 2017, the team finds a multi-dimensional universe in brain networks, which could lead to significant implications in our understanding of the brain.
• 2018, the release of the first 3D model of the mouse brain, named Atlas. The team releases Blue Brain Nexus, which would be used as a knowledge graph, for searching, integrating, and tracking large-scale data and models.
• 2019, the Blue Brain starts answering difficult questions. How does the brain find order in the sea of noise and chaos? By advanced simulations, the team found how neurons interact with each other to find even the weakest signals.
• 2020, they extend the performance modeling techniques of the brain tissue simulations. They make analytical performance models on the brain tissue simulations, which hence prove the feasibility of full brain tissue simulations.

Current stage

Till now, the team has already mapped the whole brain of a mouse into a 3D model. They also scaled up to build the smaller blocks of the Human Brain, like the Neocortex and the Thalamus. The team has developed models for large-scale visualizations as well as the methodology for saving and manipulating these maps.

Now, what helps them to make these things? The course of making went from building small, more generalized tools for generating blocks of the neocortical column. Once that was achieved, they will scale it up and predicting some of the structures. How can you predict the parts of the brain and just guess how they look? The thing is that our brain packs a highly organized network structure. And what happens when something follows a pattern? It becomes predictable. Hence scaling up from smaller structures and predicting some parts of it was possible, this is what they showed in 2015.

So, it’s not a guess. It’s a logical guess.

Tech-stack

Here are the tools used or developed in this project, that you may use to study or even make your own visualizations.

• First comes Brayns. It helps to see where the neurons fire and how they interact with each other. Brayns is made to showcase beautiful visualizations using light, shadow, and depth of field effects, to make it easier to track and understand the transfer of impulses from one neuron to another.

BlueBrain/Brayns

• But how can we generate impulses in a virtual model? How does it take the input? Well for that they use an open-source tool named, Brion. It is a collection of various input/output functions for accessing the lower-level data model, i.e. accessing the neurons and firing them.

BlueBrain/Brion

• Okay, that’s great. We have heard and also read that neurons fire specifically, and their intensity will be different for each movement. How do you distinguish one neuron fire from another? Well for that we need to extract the amplitude of the fire, which can be done by the feature extracting library named, Electrophys Feature Extraction Library (eFEL).

BlueBrain/eFEL

• Wait a minute. Do we just play with it in the instance? Can’t we save the information and maybe also search for specific neuron pathways and experiment? Well yes. For that, we are going to use the Blue Brain Nexus, which is a knowledge graph for data-driven science. Nexus can be used in any data-driven scientific field. Here we use it for what we wanted: To save and retrieve neuron pathways and other types of information stored and analyzed from the virtual brain of ours.

BlueBrain/nexus

Just one more thing, how can we showcase this in real life, as in the final output in normal form, and not computer codes and commands. Well, there comes Tide, not the detergent. Tide is a library developed to save the models and represent them on large screens and high-resolution videos.

BlueBrain/Tide

The project is huge. This is so much to keep track of. And everything at the same time feels a bit difficult to take in. Noah Hutton, a documentary maker, thought just the same. But he said to himself, we all love movies, right?

So, he went to the project lead, Henry Markram and asked for his permission to document everything they do and interview the scientists over 10 years (now 15) to make a film about it. He agreed. Since then, from 2005, Noah visits their facility in Geneva every year and documents the project’s progress and interviews scientists. He took the documentary and gave three stages to it.

• First, the start, Markram’s Vision. Talking about the goals and ambitions of this Israeli scientist.
• The second, Criticism. The Blue Brain Project has received a lot of criticism for itself and also for their leader. In 2013, the team of over 800 scientists even signed an appeal for the change in leadership. Not just that, people like Sebastian Seung has gone head to head with Henry about its whole concept and promise of virtual functional brain mapping.
• Then the third part, Competition. Many other projects other than the Blue Brain Project were founded during this decade long journey, like the BRAIN initiative by Obama and many others. The documentary is set to be released in 2020.

What’s next?

In the future, the Blue Brain Project will try to answer even more fundamental questions like the state of consciousness. How do we think? Questions like these may be answered once this project gets concluded. But where does it stop? When will we know that they did it? The team aims to create a digital reconstruction of the human brain by reverse-engineering mammalian brain circuitry. And once they do that it will be one of the greatest achievements in the field of Neuroscience. That’s when we will know.

Further reading and references-

• Labocine. (2016, December 8). Selections From Bluebrain: A 10-Year Neuroscience Documentary. Retrieved from https://medium.com/labocine/selections-from-bluebrain-a-10-year-neuroscience-documentary-56d869599640
• Blue Brain Portal. (n.d.). Retrieved from https://portal.bluebrain.epfl.ch/
• The Blue Brain Project. (n.d.). Retrieved from https://github.com/BlueBrain

Editorial note-

This article was conceptualized by Aditya Vivek Thota and written by Dishant Parikh of The Research Nest.

Go through the hyperlinks throughout the article to explore more about this project.

Do follow our publication for diverse research trends and insights from across the world in science and technology, with a prime focus on artificial intelligence!