Aug 13 2009

Vision and How the Brain Organizes Itself

Knowledge about brain organization and development is rapidly evolving, partly due to new tools we have for research, such as various MRI techniques. One type of question being investigated is the degree to which brain organization is pre-programmed in the genes vs a result of the process of development.

We know that both processes are involved. Human brains are all grossly the same – I have looked at hundreds of MRI scans of patients and they are all pretty much laid out the same way. Neurologists get very good at predicting exactly where a lesion will be in the brain based upon a patient’s symptoms and exam. This ability to localize a lesion is based upon the assumption that all brains are organized the same, and this assumption turns out to be almost always correct, at least to a level of precision required for a clinical assessment. (There are known variations in terms of which side of the brain will house certain specific functions – for example most people are dominant for language in the left hemisphere, but some left-handed people have language in the right hemisphere.) This suggests that genetics plays a dominant role in determining the large-scale structure of the brain.

But neuroscience researchers often operate at a much finer level of detail in terms of neuroanatomy, often dealing with the micro-structure of the brain. And the question remains, at fine levels of detail, how much of the brain’s structure is pre-determined by genetics, how much results from the process of development, and how much is plastic (determined by later use)?

We know that mapping plays an important role in brain development. This is a process by which neurons will map to sensory input or other feedback. When babies see, for example, their visual cortex will map to the stream of information coming in from the retina. The visual cortex therefore creates a literal map of the retina. Genes do not need to lay out this map, like a blue-print, but only have to set up the rules and then allow the mapping to take place.

Neuroanatomy of Vision

But the organization of vision in the brain is much more complex than a simple map of the retina. Visual information undergoes a great deal of processing to make the information more useful. Edges and contrast are enhanced, for example. Also information is processed in such a way as to create three-dimensionality and depth. Generally speaking, optical illusions are tricks that expose these kinds of visual processing.

But even more than optical processing, what we see also conveys meaning (arguably the ultimate evolutionary utlity of vision). The most obvious is probably pattern recognition: we don’t just see shapes, color, and depth – we see things. The famous eponymous essay in the book The Man Who Mistook His Wife for a Hat by Oliver Sacks desribes a man who had a stroke in the occipital-temporal region on boths sides. As a result he had visual agnosia – he could see but could not put all the pieces together and recognize specific things. So he could see the shape of his wife’s head, which is similar to the shape of his hat, but could not tell the two apart.

There are also specific forms of visual agnosia, or type-specific. Prosopagnosia (lesion of the right fusiform gyrus), for example, results in the inability to recognize faces. Landmark agnosia (right parahippocampal gyrus) results in the inability to identify specific buildings or landmarks.

ventral streamTo back up a bit further – our current models of visual organization in the brain is that there are two streams of visual information: the ventral stream relates to object recognition (and lesions of which result in visual agnosia, as described above), and the dorsal stream is for planning actions. Lesions of the dorsal stream result in what is called optic ataxia.

The dorsal stream of visual information is therefore related to seeing an object heading in your direction and getting out of the way, or to the act of catching a baseball. While the ventral stream will tell you that what you are catching is a baseball and not a rock.

To add a further layer of complexity, the ventral stream is divided into various regions with different specialties. The medial structures in the ventral stream relate to non-living objects, while more lateral structures specifically relate to living objects. There are also identifiable specialized regions, such as for visual word recognition and texture vs form.

But clearly, living vs non-living is an important distinction to make, and it is interesting that the visual pathways would separate out based upon such an abstract concept.

Now Back to Development

This all leads back to the original topic of this post, which is to discuss some new research relating to the development of separate visual pathways in the ventral stream for living vs non-living objects. The question at hand is whether or not this level of brain organization within the ventral stream is determined by genetics or by development and use.

To answer this question Dr. Mahon and colleagues at the Center for Mind/Brain Sciences (CIMeC) at the University of Trento, Italy, and Harvard University looked at the brain of subjects with normal vision and subjects who have been blind since birth. The assumption is that those who have been blind since birth have never had any sensory input to their visual cortex, and therefore no developmental mapping or plasticity with use has taken place. Therefore, whatever organization they find is likely innate and determined by genetics.

From the press release:

“Using functional magnetic resonance imaging, we found that the same regions of the ventral stream that show category preferences for nonliving stimuli and animals in sighted adults, show the same category preferences in adults who are blind since birth,” explains senior study author Dr. Alfonso Caramazza from the CIMeC and Harvard University. “Our findings suggest that the organization of the ventral stream innately anticipates the different types of computations that must be carried out over objects from different conceptual domains.”

Cool. So our brains evolved to inherently distinguish between living and non-living things – this ability is built into the basic structure of our visual cortex.

This reminds me of a book I recently read called Supersense by Bruce Hood, in which he argues that we have an innate sense that there is something fundamentally different between things that are alive and things that are not alive. This results, he argues, in supernatural beliefs in a life force or life essence to explain our strong innate feelings.

It seems that this latest research supports Hood’s premise – the distinction between living and inanimate seems to be fundamental to human brain organization. And if the evolutionary pressures were strong enough to result in these categories of objects being distinct in our vision, it is likely that the same is true about how we emotionally and intellectually categorize the world too.

In fact Hood goes over many psychological experiments that show just that – even infants think very differently about living and non-living things.


This is a fascinating area of neuroscience research, which is really taking off due to fMRI and other techniques. It is interesting, but not surprising, that at this fundamental level of  brain anatomy development appears to be following a genetic blue-print, rather than just reacting to input or use.

9 responses so far

9 Responses to “Vision and How the Brain Organizes Itself”

  1. artfulDon 13 Aug 2009 at 12:49 pm

    Living things can eat you, especially if you fail to distinguish them from knick knacks. Actually the distinction is about the potential that an object has for predictable behaviors.

  2. Enzoon 13 Aug 2009 at 12:57 pm

    Patterning is amazing.

    I am not very surprised to find that our brains are patterned at least partially by genetics, but I’m still astounded that the entire process of development occurs with exquisite reproducibility from individual to individual. There are roughly 30,000 genes in the human genome, hardly enough to encode for a detailed map of every organ system’s structure, especially considering the complexity of the brain. There is no gene that says “put this cell here juxtaposed to this cell and form this type of connection between them.”

    It’s just so cool that organisms have developed to code for the building blocks and leave the rest of the building up to the laws of chemistry and physics.

    Thanks for the insight into the brain. I think a lot of us underestimate how much we are learning about the biological frontier.

  3. twanvlon 13 Aug 2009 at 1:04 pm

    One question that immediately comes to mind is: what is considered living? Do we recognize robots as alive? How about dolls?

    If the specific instances are learned, then how does the brain make a choice of which region to use? Perhaps based on whether it moves unpredictably? Or on an animal-like shape?

  4. daedalus2uon 13 Aug 2009 at 3:50 pm

    One of the best documented cases for NO being important in neurodevelopment is its role in the refinement of the mapping fidelity between the retina and the visual cortex. If nitric oxide synthase is inhibited the fidelity of that mapping is poor. This refinement occurs largely by the elimination of connections under conditions where nitric oxide synthase is not inhibited. Presumably the importance of NO in refinement of connections is not limited to the visual cortex.

    In the fMRI BOLD technique, where changes in the levels of oxyhemoglobin are measured in real time by how the magnetic susceptibility affects the MRI signal, what causes the differential blood flow is acute changes in the level of nitric oxide. Levels of NO in this range do affect long term potentiation and long term depression, so presumably they are important in neuronal remodeling even in adults.

  5. Woodyon 13 Aug 2009 at 3:50 pm

    Very interesting. I am curious whether the ventral stream partitioning is truly based on living vs. nonliving or animate vs. inanimate. In other words, would a tree or a barnacle activate the medial or lateral ventral stream?

  6. Yngveon 16 Aug 2009 at 3:17 pm

    Interresting stuff.
    All of this ties neatly into what I’ve read, from Steven Pinker, Bruce M. Hood, Gary Marcus, David J Linden and, ofcource, this blog (been lurking for some time now).
    And it makes perfect sense.
    I’m really excited as to what future research will show. Are there more innate distinctions lurking in our head? (probably)
    – And what will they tell us about ourselves?
    Is there plasticity in other areas that may tie into innate functions, making a difference in the pronounciation of their effect?
    Sort of piggybacking on an existing feature, promoting one or more aspect whilst supressing others…
    (Did those last sentences even make sense…? I’m not a neurologist, hehe)

  7. stompsfrogson 19 Aug 2009 at 4:51 pm

    I like fingerpaints.

  8. artfulDon 19 Aug 2009 at 6:08 pm

    Stomps, that was one of the best references ever. Tends to confirm we are born with the capacity to model what we can expect our sensory apparatus to discover, and to do so with regard to size, distance and perspective – which would certainly help in assessing the potential for any dangers or rewards involved.
    As to colors, it seems clear that we not only expect to see them, but based on the way the artist seems to match colors with natural objects, we could have already attached some meaning to those observations.
    Of course somebody has to have helped with pigment selection and thus may have told this artist what to put where, but there has been other evidence that colors have inherent meaning to us, and this could help to confirm it.

  9. artfulDon 19 Aug 2009 at 8:52 pm

    For what it’s worth, this is the blind artist’s website:
    And here are some of the color paintings:

Trackback URI | Comments RSS

Leave a Reply

You must be logged in to post a comment.