Connectome

The revolution of brain imaging and the general burst of technology meant new possibilities, and the adaptation of the field to those new possibilities.

As impressive as it is, neuroimage has an unsolved problem. The macroscopic approximation of fMRI loses cellular detail. The basic unit of 3D images, the voxel, has a 2mm3 , space that can contain thousands of cells with millions of synapsis. In this scale, is impossible to detect activity events, like action potentials or neurotransmission between neurons.

This brings us to the present, where the new objective, is to obtain a complete description of the brain, from its basic components to the complex dynamic mechanisms behind. Connections and activity in the structures.

The revolution of brain imaging is the hot topic of the XXIst century, and in the present a lot is going on. We are confronting and advancing in the challenge of mapping the brain.

Connectomics

Objective: to obtain a detailed map of the exact trajectory of every axon in the brain, and to decipher the complex net of connections composed of infinite possibilities.

In the decade of the 70s, Sydney Brenner accepted the challenge of mapping every connection in the worm, Caenorhabditis elegans. This is the technical foundation of Connectomics.

In spite of this, we own the term to Olaf Sporns and Patric Hagmann. They used this term independently, but simultaneously, and is based in a version of the word, genome.

The challenge proposed is to obtain a tridimensional and ultrastructural map of the brain trough neuroimaging techniques.

Caenorhabditis elegans is an organism with 300 neurons and around 7000 synapses. Brenner and his team dedicated a decade to finalize the counting of the entire connectome of this worm.  The project finished in 1986.

The implemented technique is not valid for the human brain, because it would take millions of years.

There are other projects that follow different counting techniques, like Brain Connectivity, directed by Allen Mouse.

Allen and his team, have completed a map of the neuronal projections of the mouse using advanced optic microscopy and special staining methods, tract markers that only identify projections. Is more rapid than Brenner’s method.

The human connectome project (HCP), born in 2009, has as main objective, to trace a map of the connectivity of the human brain, based on a group of 1200 subjects.

For the time, to make the mission, even possible, they work with macroscale. Instead of tracing individual neurons, they trace regional connectomes.
This regional connectome is extracted from the visualization of the brain activity, provided by fMRI and diffusion magnetic resonance imaging (IRMd). 


Fig 1. Courtesy of the Laboratory of Neuro Imaging and Martinos Center for Biomedical Imaging, Consortium of the Human Connectome Project – www.humanconnectomeproject.org

iIn the end, even if it’s true that knowledge, and a great quantity of data about general patterns of connections can be obtained with these last inventions, it’s not possible to identify the individual connections of a neuron. Therefore, a detailed connectome of the brain is not realistic with the actual resources.  

Investigations in genomics and high spatial resolution microscopy are being carried out in order to solve this problem.

So far, the closest advance in the matter, has being tackled by Winfriend Denk, who in 2004, introduced the Serial block-face scanning electron microscopy (SBFSEM), a microscope that includes a ultramicrotome in the vacuum chamber.
This microscope generates high resolution three-dimensional images from small samples.

The  SBFSEM was first used to analize the connectivity of axons in the brain. It has enough resolution to trace the smaller axons and synapses. Even if it’s being perfected, it can already mange a large quantity of data and implats algorithms in order to apply segmentation.

The eyewire project, by Sebastian Seung, is presented as an online game, and it uses images obtained with an SBFSEM. The participants have to search for neurons in the images presented, and at the same time the collaborate.

Cartography of neural activity

Objective: to describe the bioelectrical traffic flowing trough the paths of the connectome.

To elucidate the connections in the brain is important, but to fully understand how it works is necessary to add neural activity in the equation.

Neurons that fire together wire together.”

Donald Hebb, 1949

This quote summarizes the essence of neural activity. To add some complexity to it, we can say that is also important to know the molecular mechanisms causing those shooting sequences and their modulation.

This thought carry us to the dynamic functioning of the brain, in a spaciotemporal conception.
In this case, the challenge is to identify each one of the elements of neural activity taking place during specific cognitive processes.

Brain, the project conceived by Rafael Yuste and founded in 2013, attempts to investigate the brain trough the advance of innovative technologies. They use optic electronics and image techniques, but they also develop new ones using advances in syntetic biology and nanotechnology.

One line of investigation, is an effort to create nanowaves that could act as sensors to register the activity of individual neurons.
They also aply IRMf and calcium imaging to capture cerebral activity in real time.

A member of this project, George Church, started “Rosetta Brain” an initiative that applies Fluorescent In-Situ Sequencing of Barcoded Individual Neuronal Connections (FISSEQ-BOINC), a genetic strategy to reveal the mapping of every neuron in the brain in rich molecular detail.

Usual tecniques in the tracking of neural activity

The EEG technique invented in 1929 by Hans Berger, is an excellent tracker of neural activity regarding temporal resolution. It gives out a registry of electrical signals in the brain coming from specific  neurons, and it provides with the oscillatory waves from the sincronization of populations of this nervous cells.

In the present, the original electrodes have been substituted for high density multi-electrodes and multielectrode arrays, because they affine the registry of the physiological signals.  The problem of this registry is the poor spatial resolution. The exact procedence of the signal is unknown.

Calcium imaging

Junichi Nakai, invented this technique in 2001, where changes in the fluorescence of a molecule he created, GCaMP, related to the level of calcium concentration, can be measured with a fluorescence microscope.

All of these efforts are still not suffient to obtain the desired global image of the brain. The challenge of building a map that incorporates brain activity, the connectome and other relevant elements, continues.

The project Mindscope, started in 2013 in the Allen Institute for Brain Science. This project aims to reunite data on activity, connectivity, morphology and molecular characterization of the neurons to describe, profoundly, the existing types of neurons populating the cerebral subsystems. In the present they are focused on the visual system of mice.

Big scale simulation of the brain

Objective: realistic biophysic models capable of producing the computations and algorithms that the brain generates in order to explain mental functions.

Even if it’s possible to obtain all the data that the projects mentioned above aim to obtain, the following challengue would be to find the way to manage and process that amount of information.

This bring us to the last part of the debate, the in silico representations.

When trying to stablish a casual relation between the mapping of the connectome and the mapping of the brain activity, the most intuitive possibility and the most plausible way of connecting this two aspects, is the simulation.

Even if it’s suspected that mental functions are a product of brain circuits, the truth is that there is no irrefutable proof of this. So we are back where we started.

How, a concrete, detailed pattern of molecular changes, synaptic events, and action potentials explains a mental process?

If we answer this question, it has to be trough a model. An integration, of every process happening in each neuron at a certain moment in a particular context.

A mathematical model is an abstraction that relates the components of a system with the events that generates them.

There is a principle of computation in the brain at the level of individual neurons. From thousand of synaptic contacts, and a particular excitability, each neuron generates, or doesn’t generate an action potential. Yes or No, 0 or 1.

The problem is, the input, that each neuron recieves is extremely large.

Some models

Mindscope, is trying to create a model from the collective dynamic of big nets of neurons, their variations during behavior.

The Human Brain Project, an EU initiative, is trying to integrate data from numerous investigations in order to integrate them and make a simulation according to them. It’s a platform open to investigators that allows the creation of models in every level of complexity.

Spaun, is a model of the human brain created by a group of neuroscientists in the University of Waterloo. This machine, running in a supercomputer counts with 2.5 million simulated neurons that are included in simulated brain regions. It also has a digital eye, and a robotic arm. It is trained to perform 8 tasks, from drawing to fluid reasoning.

Fig 2. Artificial inteligence model, Spaun.

References

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this:
close-alt close collapse comment ellipsis expand gallery heart lock menu next pinned previous reply search share star