Who said that research does not pay? That this is not the case is
demonstrated by the Human Genome Project, launched by the U.S. in 1999 at a cost of slightly
less than $ 4 billion in 10 years but that, to date, has yielded $ 796
billions in business activities and has created 310,000
new jobs.
Now it is the turn of the
Brain Activity Mapping Project, for which Obama is going to spend a little
more than $ 3 billion over 10 years and whose official announcement is expected
this month (March 2013). The Brain Activity Mapping Project is the U.S. counterpart
of the European Human Brain Project, which will be funded with 1.2
billion dollars over 10 years.
Beyond the common theme, the brain, the two projects are
however very different, both in terms of methodological approach and as to their purpose.
The Human Brain Project objective is to build an artificial brain capable of
performing some of the human brain functions
using ICT tools (Information Communication Technology). However, it is known that the same result
can be obtained through completely different algorithms. Therefore, the fact of
performing in silico some of the human brain's
operating functions does not provide an assurance that this is done using
the same computational processes used by man.
Unlike
the Human Brain Project, the Brain Activity Mapping Project intends to explore the neural processes as they occur in
nature, through a study of the functional activity of all the elements
that constitute the brain. This learning process will proceed in
steps, starting from the nervous system of simple organisms, such as the worm Caenorhabditis
elegans, the fruit fly Drosophila melanogaster and the
zebrafish Danio rerio, up to the primates and humans, passing through the
laboratory mouse Mus musculus and other simpler mammals.
The study should be based on the method that
since the inception of neurosciences has been the method par excellence to study
the function of the nervous system, i.e. the recording of the neurons'
functional activity. In this case, however, the scale of the study is different as it is no longer based,
as it currently is, on recording a sample of a few representative
neurons, but on the simultaneous recording of the activity of all the
neurons that make up a circuit. The human brain has about 90
billion neurons, cells that, unlike those of other organs (the
liver, muscles, mucous membranes, etc...) show extreme heterogeneity and
specialization and are connected to each other through specialized
membranes (e.g. synapses), which further amplify the complexity.
Given that, the Brain Activity Mapping Project is not
based on new working hypothesis (hypothesis-driven), but on the development of
new and revolutionary methods (methodology-driven). If we have to define
the core concept underlying this project, we could only say that it is
the brain study methodology that must conform to its complexity.
The project will involve the development of recording methods
that must be equipped with a space-time resolution
adapted to the characteristics of individual neurons' activity, while at the same
time being capable of simultaneously recording from hundreds of thousands of them.
There is no doubt that the aim of the project is very ambitious and
the result cannot be taken for granted. For this reason, the project has been questioned by
various opponents before it even commenced, including neuroscientists, on the basis of a number of arguments.
The most serious one is that the methods capable of recording from populations
made up of thousands of neurons do not possess a sufficient spatio-temporal resolution
to record the activity of the individual neurons composing them. On the other hand, methods that allow for the
recording of individual neurons activity, including those using
optical methods, are invasive and therefore have limited applicability
to humans. Finally, it has been objected that the recording from such a
huge multitude of neurons would lead to major difficulties in
data storage. For example, in order to store 10 million
neurons per year would require 30 petabytes of storage space.
The super-accelerator (LHC) in Geneva each year generates data that occupy about 10 petabytes. The data generated by recording the
activity of 90 billion neurons in the human brain would occupy a space
of 300,000 petabytes per year.
In principle, the Brain Activity Mapping Project should give a major impetus to the Neurosciences. However, these large projects do not necessarily have positive effects on research in general. For example, the fact that the LHC has accurately confirmed the Standard Model, in the end has caused scientists to desert a very fruitful area in theoretical physics, that of the development of alternative models to the Standard model. The Brain Activity Mapping Project may in turn absorb the resources available for the neurosciences, already reduced due to cuts in public spending.