Neuroscience

How do concepts of the very large and potentially the infinite arise in the study of neurobiology?   Our first contributor in this field is Stuart Thompson,  professor and neurobiologist at Stanford University.  With his kind permission, the following is excerpted (minus images) from his fascinating blog – Neuromavin. http://neuromavin.com.

The Human Brain and Other Really Big Things

Some systems can only be described by very large numbers. Take the Universe, for example. How many galaxies are in the known universe? The best estimate comes from the Hubble Extreme Deep Field study which suggests somewhere between 100 and 200 billion, although other simulations put the number closer to 500 billion. These are big numbers; in scientific notation they are 1 – 5 x 1011 (1-5 followed by 11 zeros). But if we go further and consider the number of stars in all those galaxies, we get to a very healthy, very big number. Assume each galaxy has one hundred billion stars (1×1011; a conservative guess), then the number of stars in the known universe is possibly 1×1022, and that is without even considering multiverses. What can compare?

Computer engineers deal with large systems, how large?   IBM is building a computer based on its SyNAPSE chip. The device uses a network of elements, termed neurosynaptic cores, connected in a fashion loosely based on the long-distance wiring of the monkey brain and with cores clustered into brain-inspired regions.  Ray Kurzweil, currently at Google, summarized IBM’s progress to date (http://bit.ly/1Ly5rgy). Teaming up with Lawrence Livermore Lab, IBM achieved a prototype emulating 53×10¹⁰ neurons and 1.37×10¹⁴ synapses (~100 trillion).

Thinking about supercomputers set me to wonder how big is the human brain? This is not a simple question because there are so many bases on which to make comparisons. Volume is out because computers are just enormous hulks. One has to ask, what is the most meaningful unit of measure? For systems that deal in information, it might be the number of processing units and the connections between them. It also might have something to do with the fundamental mechanisms used to handle information, be they electrical, optical, chemical or other. There are other meaningful units, I am sure, but let’s take these as a starting point.

All nervous systems have a cellular design and information is transferred between cells at synapses. So how many synapses are there in the human central nervous system? Good estimates range between 100 trillion and 1,000 trillion; expressed in scientific notion the number is 1X 10^14-15. A very large number of connections, on a par with IBM’s finest. This makes it clear that the human brain is a complicated system in the sense that a supercomputer or a universe is complicated, and large on a similar scale. I would argue, however, that the brain is a complex system, as well, and I do not think that the number of synaptic connections give a meaningful representation of information processing power. We have to add some more dimensions because the brain does not run out of tricks when it runs out of synapses and the most important is probably the ability to change, i.e. plasticity. Synapses can be added and deleted, they can change strength in response to experience and new connections can be made over short and long distances. This is the stuff of learning. New neurons can be added by the activation of stem cells, and in at least one part of the brain, the hippocampus, this happens throughout life. That is part of learning, too. All of this plasticity occurs on a very rapid time scale, sometimes overnight; a rate that leaves the cosmos and Moore’s Law is the dust.

We have talked of three large things, the universe, a supercomputer and the human brain. All three are complicated because they have an enormous number of parts. In addition to being large, the brain is also complex because it is self-assembling and because of its tunable elements, capable of changing dynamically in response to events both real and imagined. That is the beauty and the advantage of biological systems. If you are interested in big numbers read this mind spinning discussion by Alasdair Wilkins http://bit.ly/1hoAKJi.