Monday 2 May 2011

Networks in the brain

So your friend Ben is married to Margaret who is friends with Rachel who shares an office with Rory who worked on a planetarium show with Rob who once received a detention at school for mooning Prince Harry [*]. 

According to the theory of the six degrees of separation, you are no more than half a dozen people away from receiving that front row invite to the Royal Wedding. The idea is that you are connected to every other person on Earth through an average of six people. It is a concept huge social network sites such as Facebook have been testing, but surprisingly it is an arrangement that is reflected in the structure of your brain.

All this I learned at the Royal Canadian Institute (RCI) 2011 Gala. The RCI was formed in 1849 by Sir Sandford Fleming. One of its original roles was to publish a scientific research journal in Canada but now its emphasise is on a weekly public lecture series which covers a wide range of scientific topics. In addition, the RCI helps with grants for students wishing to study science at university and it hosts an annual Gala dinner. The Gala is an opportunity to have a discussion over a great meal with a scientist. One of the twenty five tables at this year's event was hosted by my adviser, Professor Ralph Pudritz, but I shunned his table in favour for one led by a scientist working on the structure of the brain; a topic I knew nothing about. (When I told Ralph I'd rejected his table in favour of another he assured me he 'expected nothing less'. I don't think he meant this to be a reflection of my attention in our research group meetings.)

Our table was led by Professor Mark Daley who worked on models of the brain at the University of Western Ontario. When newly arrived at his institute, Mark explained that he had known very few people.

"But, I did know Mike." He gestured towards one of the other diners seated with us. "And Mike knew everybody. So if I needed to contact somebody elsewhere in the University, I could go to Mike and the chances were he knew them. This meant although I only knew a few people, I was connected to almost everyone else via only one person."

This, Mark explained, was the premise behind the six degrees of separation. There are a few people who know a huge number of others and these individuals act like hubs. People preferentially attach themselves to hubs (since the hub is likely to meet them through their enormous list of contacts) resulting in them being connected to a great many others through a very small number of steps.

What Mark said about Mike turned out to be entirely true. When chatting to him before dinner he had declared, "Oh, you're at McMaster! Do you know Hugh Couchman and James Wadsley?" I had to confess I did.

Mark continued by explaining that the brain organises its neurons along similar principals. There are hub areas in the brain which have a huge number of neurons connected to them and these link up regions which have sparsely few connections.

This structure can be explored with two major methods. The first is to take thin slices of the brain's grey matter and the second (more desirous for live volunteers) is to watch water flows via an MRI scan.

The consequences of this neural structure have important ramifications both for the effect of brain-damage and in understanding mental illnesses. Damage to one of these hub region, for instance, can result in the head injury being fatal because the brain simply cannot rewire to compensate from such a large loss of connections. Other times, the damage can be severe but limited to one specific area. Mark cited an example of a woman with damage to one hub who was left unable to see.

In most people, the number of hub regions is small and they are found is quite specific areas. One exception to this is in the case of people suffering from schizophrenia, where many smaller hub nodes are seen and in farther flung areas in the brain than for a healthy person. 

A question I asked was whether this was the underlying concept in electric shock treatment for depression? Was the idea to try and forcibly rewire the neurons by destroying their electrical signals and thereby forcing the brain to choose another (hopefully better) structure? Mark said that while this was the correct premise, such treatments were now strongly out of favour. He compared it with chemotherapy, saying you effectively killed a lot of neurons in the hope that you destroyed the bad pathways before you took out all the good. He did describe less invasive treatments which included asking the patient to think of something pleasurable directly after thinking of a traumatic event. Over time, the association can force the brain to rewire and help with post-traumatic stress disorder.

So what is is that governs our thoughts? Is the brain, as Penrose claims in 'The Emperor's New Mind', a system governed by random probabilities via quantum mechanics? Or are we, as Mark assumes in his work, simple Turing machines whose thoughts and actions can be completely predicted based on our experiences? Neither sounded particularly appealing.

"I want another option," I told Mark. He nodded and promised me one after he'd finished his dinner. The problem with being the guest speaker at a meal was the actual food was hard to fit in amidst the barrage of questions.

The third option, he explained as the plates were cleared, was that our mind is like a Bayesian machine which using a mixture of probabilities and input from its surroundings to make decisions. So when faced with the delectable crumble for desert, there was a very high chance that I would take the logical choice and eat it. Then there was the small probability I'd lob it across the table. I love feeling I have choice.

The crumble was rhubarb, in case anyone was wondering.

At the end of the dinner, each table was allowed to pose a question to another group to allow diners the chance to hear about the different areas being discussed that evening. The most important question was posed first and was directed at Professor Jeffrey Rosenthal from the department of statistics at the University of Toronto:

"What is the probability that Kate Middleton will wear a slinky wedding dress?"

"Slinky?" Jeffrey rose to answer the question. "This isn't as close to my area of expertise as you were led to believe!"


--
[*] Editor's note: any resemblance to real people, in the Physics and Astronomy Department or otherwise, is purely coincidental and Rob has never yet admitted to knowing Prince Harry. Or mooning.

No comments:

Post a Comment