Field of Science

  • in The Biology Files
  • in inkfish
  • in Life of a Lab Rat
  • in The Greenhouse
  • in PLEKTIX
  • in Chinleana
  • in RRResearch
  • in The Culture of Chemistry
  • in Disease Prone
  • in The Phytophactor
  • in The Astronomist
  • in Epiphenom
  • in Sex, Genes & Evolution
  • in Skeptic Wonder
  • in The Large Picture Blog
  • in Memoirs of a Defective Brain
  • in C6-H12-O6
  • in The View from a Microbiologist
  • in Labs
  • in Doc Madhattan
  • in The Allotrope
  • in The Curious Wavefunction
  • in A is for Aspirin
  • in Variety of Life
  • in Pleiotropy
  • in Catalogue of Organisms
  • in Rule of 6ix
  • in Genomics, Evolution, and Pseudoscience
  • in History of Geology
  • in Moss Plants and More
  • in Protein Evolution and Other Musings
  • in Games with Words
  • in Angry by Choice

A Cubic Millimeter of Your Brain

Are there more connections in a cubic millimeter of your brain than there are stars in the Milky Way? We are going to answer that question in a moment, but first take a look at this image of hippocampal neurons in a mouse's brain. It is an actual color image from a transgenic mouse in which fluorescent protein variations are expressed quasi-randomly in different neurons. This kind of image is known as a brainbow and is aesthetically awesome further it may be one way to empirically examine a cubic millimeter of the brain (neuron tomography).
Hippocampus brainbow
by Tamily Weissman, Harvard University
In reality mapping even an entire cubic millimeter of the brain is an extremely daunting task, but we can still answer my original question. First, I know that there are different kinds of neurons that vary in size and that some neurons can have a soma (the big part that has the nucleus from which the dendrites extend) spanning a millimeter in size. Thus if you picked a random cubic millimeter of brain you could run right into the heart of a neuron and you would find very few connections. Given this fact, we can very easily answer this question with a resounding no, however, this seems like an unsatisfactory trite approach. So I looked up some numbers on how many neurons are in the brain, how many connections are in the brain, and how many stars are in the Milky Way. Lets answer the question using the 'average' number of connections per cubic millimeter.

How many neurons and connections there are in the brain? This is kind of a tricky question and I am not a nuerobiologist so I have gone to several resources for the answer. Professor of Computational Neuroscience at MIT Sebastung Seung says in a TED talk
your brain contains 100 billion neurons and 10,000 times as many connections
Professor of Molecular Cellular Physiology at Stanford Stephen Smith says in a press release on brain imaging that
In a human, there are more than 125 trillion synapses just in the cerebral cortex alone
René Marois from the Center for Integrative and Cognitive Neurosciences at Vanderbilt Vision Research Center states in a recent paper [1]
The human brain is heralded for its staggering complexity and processing capacity: its hundred billion neurons and several hundred trillion synaptic connections can process and exchange prodigious amounts of information over a distributed neural network in the matter of milliseconds.
I have enough expert sources now to confidently say these experiments agree that the human brain has some 100 billion neurons (1011). The number of connections seems less precise, but it is at least several 100 trillion connections (1014) as judged by Marios and Smith and as much as 1015 as judged by Seung.

The number of connections in the brain is tricky to define. We may define a synaptic connection as each place the neuron touches another neuron and a synapse is present. It doesn't seem to make sense to simply count incidental contact. Further, there is the question of whether we should count redundant contacts between neurons. We can obtain an upper bound on the number of connections in the brain by considering the case in which every neuron is connected to every other neuron. Coincidentally the operation of connecting every node in a network with every other node is a process I am familiar with from cross correlating radio signals. Anyways, the equation we are looking for is N(N-1)/2 where N is the number of nodes in the network. Thus, for our N=1011 neurons the maximum number of non-redundant connections is about 1022. This maximum bound is huge! But how huge is it really? Hilariously, while searching for an answer to my original question I found a message board pondering the grand statement
There are more connections in the brain than atoms in the Universe.
A really clever person pointed out that
Theoretically, if we took all the atoms in the universe; wouldn't that include the atoms within the brain?
People have this feeling that the number of connections between items can be much larger than the number of actual items in the collection and while this intuition is true the idea that there are more connections in the brain than there are atoms in the universe is absurd. Lets put it in perspective that a few grams of any substance, like water, is measured units of moles. A mole is standard unit of measurement corresponding to the absolute 6.02 x 1023. Thus even a drop of water contains more atoms than there are connections in the brain.

Now we need to know how many neurons and connections are in an average cubic millimeter of the brain. How big is the brain? John S. Allen of the Department of Neurology at University of Iowa stated in a recent paper that[2]
The mean total brain volumes found here (1,273.6 cc for men, and 1,131.1 cc for women) are very comparable to the results from other high-resolution MRI-volumetric studies.
We can take the volume of the brain as 1000cc as a low estimate (which will only over estimate the density of connections).

The final thing we need to know to answer the question at hand is the number of stars in the Milky Way. Like every other number we have been working with it is rather uncertain. Even if we define a star as only those spheres of gas which are large enough to fuse hydrogen at some point in their lifetime we don't know the answer because we can't see the multitudes of dim stars. There are probably at least 500 billion star like objects in the Milky Way. Lets take 100 billion as the number to be conservative.

Finally, lets bring all the numbers together. One cubic millimeter is 1/1000 of a cubic centimeter and 1/1000000 (10-6) of the entire volume of the brain. We can scale the total number of connections in the brain (using the high estimate of 1015 connections in the brain) then we find that there are 109 connections in a cubic millimeter of the brain. The 109 connections in a cubic millimeter of the brain is two orders of magnitude smaller than a low estimate of the number of stars in the Milky Way. No, on average there are not more connections in a cubic millimeter of your brain than there are stars in the Milky Way. 

My first response to this question was bullshit! This question (or rather statement) is made by David Eagleman here at a TEDx talk and here on the Colbert Report. Colbert also called out Eagleman when he dropped this factoid, but it didn't stop the interview. I have also contacted some actual neuroscientists to see what they thought of this statement and they agree with me that it is not true. Maybe there is special part of the brain particularly more dense in connections than the brain on average, but that would be misleading like saying the density of the Milky Way is that of water because, you know, certain parts of the Milky Way are water. The better statement would be to say that there are are more connections in the brain than there are stars in the Milky Way. As Colbert would say, I am putting you on notice Eagleman.

While we are on the subject I want to mention my favorite talk about the brain which mixes just the right amount of wonder and fact. It is the TED talk I mentioned earlier by Sebastian Seung on what he calls the connectome - the network of connections in your brain between neurons which physically dictates how you think. In the video he discusses another volume tomography technique in the brain using a cube of mouse brain tissue just 6 microns on a side. It is another great visualization for what is actually in a cubic millimeter of your brain.

ResearchBlogging.orgReferences


[1] Marois, R., & Ivanoff, J. (2005). Capacity limits of information processing in the brain Trends in Cognitive Sciences, 9 (6), 296-305 DOI: 10.1016/j.tics.2005.04.010

[2] Allen, J., Damasio, H., & Grabowski, T. (2002). Normal neuroanatomical variation in the human brain: An MRI-volumetric study American Journal of Physical Anthropology, 118 (4), 341-358 DOI: 10.1002/ajpa.10092

10 comments:

  1. I never knew hippocampal neurons could look so good! Interesting and a great read :)

    ReplyDelete
  2. Eagleman misquoted himself in the Colbert Report. He does state 1 cubic millimeter in the show. However in his book Incognito, Eagleman makes two claims:

    "A typical neuron makes about ten thousand connections to neighboring neurons. Given the billions of neurons, this means there are as many connections in a single cubic centimeter of brain tissue as there are stars in the Milky Way galaxy." Pg 1-2.

    He gets a bit more specific later:

    "In a single cubic millimeter of brain tissue, there are some one hundred million synaptic connections between neurons." Pg. 173

    Eagleman, David. Incognito: The Secret Lives of The Brain. New York: Pantheon Books, 2011.

    The above quotes can be physically viewed in Amazon.com's Look Inside feature:

    http://www.amazon.com/Incognito-Secret-Lives-David-Eagleman/dp/0307377334/ref=sr_1_1?ie=UTF8&qid=1311852390&sr=8-1

    ReplyDelete
  3. This is David Eagleman. First, thank you for your blog post and careful analysis. Indeed, I (embarrassingly) misspoke my units on national TV -- immediately after it came out of my mouth I wanted to make the correction, but as you can imagine, things move pretty quickly on Colbert. As the above reader correctly noted, I meant to say cubic centimeter (not millimeter), which contains ~10^11 connections. As you can verify, I have the units correct in the book. I'm hoping to retrospectively clean up my verbal error online. Thanks for your help in doing so.

    ReplyDelete
  4. Should be N(N-1)/2 for number of connections with N nodes.

    ReplyDelete
  5. A nice and slightly mindbending (hear, hear) juggle with numbers, allowing a good insight on what's behind some of the more popular sciences today (unlike my one) ;)

    ReplyDelete
  6. Dr. Eagleman, thanks for clearing that up. Author, thanks for a great post.

    ReplyDelete
  7. David Eagleman thanks the the clarification. Stated like a true scientist. And Phantom Scientist thank you for your insightful comment and discussion.

    ReplyDelete
  8. Thank you for clearing that up. Nice work. What I would like to know is how many signals are going through those neural connections in a given amount of time, say a minute?

    ReplyDelete
  9. thanks for the clarification

    ReplyDelete

Markup Key:
- <b>bold</b> = bold
- <i>italic</i> = italic
- <a href="http://www.fieldofscience.com/">FoS</a> = FoS