[entell]

I am curious about how neurons connect to each other. OnIntelligence doesn't seem to mention how connections are made. Maybe it is common knowledge, but I don't know.

Are the connections static, determined at birth or until a certain age? or can they change over time? And if they can change, how does that work, how would neurons know which other neurons to connect to if there already is no connection hence no way to predict which other neurons are "interesting"?

[Photron]

Neural connections regulation, growth etc are really not understood. I believe, how the brain and neurons come to be is very much on the cutting edge of recent research. So this is certainly not common knowledge.

Are the connections static? Due to advances in functional imaging of the brain, researchers are finding out that older brains can indeed show activation (for example see a University of Illinois at Urbana-Champaign http://www.news.uiuc.edu/news/06/0216brain.html). These and other studies show that older brains can still grow.

This brain functional imaging is teaming up with genetics in a new area of research called "imaging genetics" to show how certain genes can determine a person's brain-type (see http://www.sci-con.org/tiki-read_article.php?articleId=255). This area of research should bring some light to how certain genes can affect how neurons grow to become the brains they are.

You asked some pretty big questions. How neurons make connections in the brain and change with learning and other factors - may be answered in our life time thanks to ever more powerful brain imaging and other techniques.
_________________
Photron
Ultimate Mind

[entell]

Hi Photron,

Thanks for your reply. Now I see why everyone seems to be avoiding this particular issue. Since I am a neuroscience newbie, I had assumed it was common knowledge..

Well, we know for sure that people keep learning new things in older ages as well. Even if it turns out no new connections are made after a certain age, it is not all that important. From this perspective, it is irrelevant, but kinda interesting to know nonetheless.

As I posted in the OP, I am particularly interested in how the brain figures out which connections to make. I read that in a child's brain, it is sort of random. The ones that work out get stronger, and the meaningless ones sorta die away.. Of course, there is nothing random in nature, so I am sure there is an answer waiting to be discovered.

[chatham]

developmental approaches to computational modeling are indeed on the cutting edge of neural network research. in fact, one of the most interesting recent trends has been a move towards embodied neural network modeling of development, also known as developmental robotics, or autonomous mental development.

not much is known about the neural pruning process during childhood. however, there have been several very successful models of developmental progression using recurrent connections (to simulate growth-related changes in prefrontal cortex) and others using the cascade correlation algorithm.

i have compiled a complete list of developmental computational cognitive neuroscience papers since 2000 at http://www.citeulike.org/user/neural_nets_chapter . Feel free to browse, or check my blog, which focuses on precisely this topic.
_________________
http://scienceblogs.com/developingintelligence

[raydac]

I am sure that neurons can change their links between each other and that would be logically but I didn't have any book where I could read about that..
I have read a book of academic Behterewa where she writes that she developed a hypothesis that there are two mechanisms at a human brain - fixed links and flexible links (I hope that I am right in my english translating)
but any way, our current knowledges about tha human brain (and a brain in general) is very poor..
_________________
http://www.igormaznitsa.com