I've been catching up on Charlie Rose's brain series. The second episode is all about the tricky and fascinating world of perception.
Brief Topic Overview
Trying to determine which of our senses is most important is like asking which leg best helps us walk: I'm thankful for each of them and would rather not give up any. Something fairly more certain is that asking this question can inspire a spirited discussion.
In terms of neuroscience, a good argument can be made for the value of sight. Roughly a quarter of our brain is devoted to vision, and this hefty portion is among the best understood. A long line of neuroscience careers have achieved fame because of their discoveries in the visual system. The advances in our grasp of the visual system have informed our view of the entire brain; it is a model for how other systems function. By understanding how the brain processes vision, we can make inferences about how the brain works as a whole.
I've often felt that visual systems neuroscientists think of themselves as an elite crowd of brain scholars, studying more important questions than the average stock. I imagine them confiding in each other, "You know, if God were a neuroscientist, He would study the visual system."
To tackle this topic, Charlie Rose and co-host Eric Kandel were joined by top scholars of the brain's visual system. Occupying seats at the round table were Anthony Movshon, who studies the visual cortex of monkeys; Ted Adelson, a researcher trying to create machines that can see; Nancy Kanwisher, a leader in the study of human facial perception; and Pawan Sinha, who studies brain development and is renowned for restoring sight to children born blind due to cataracts. With their combined expertise, they discussed how the brain processes the visual world, from faces and landscapes to optical illusions.
Key Insight from the Program
The eye is not a camera
When you snap a shot of your buddies with your digital camera, the image is captured, pixel by pixel, exactly as it appears in front of you. The camera accurately reports what colors it sees, where it sees them and how bright they are. That's not how we see the world. Gestalt psychology shows that we see images as whole objects, not simply a series of lines and colors. We draw out contrasts, such as edges and shadows, to help reconstruct a three-dimensional world in our brains. You can recognize a tree at dawn, noon and dusk even though the colors look different in these three light conditions. Processing images as objects helps explain why humans are far better at recognizing faces than computers.
According to Movshon, "Humans can throw away useless information, whereas a computer can't. We can't see a flat cube. Even if it's drawn on paper, it appears three-dimensional. "
Visual computations are hierarchical
The visual system works like an assembly line. When light enters the eye, it comes as raw, unprocessed material, but at the other end it comes out as a finished product. The eye's photoreceptor cells respond to dots of light and begin organizing these dots before sending them off to the brain. In the primary visual cortex, the brain receives the dots and puts them together into contiguous lines and streams of color. These lines and colors move on to more sophisticated visual regions, where they're packaged as an object, such as a face. From there, it can be compared to faces in your memory, and if it matches your grandma, you know to say hello and give her a hug. If it's not her, you won't be surprised if you get slapped after calling her Grandma.
Much of our understanding of the visual system came from the pivotal findings that earned Hubel and Wiesel the Nobel Prize. They showed awake cats images on a screen while recording from different brain regions. Early regions in the visual stream responded any time the cat was shown a light. However, while recording from the visual cortex, they found that cells responded only to lines in specific orientations. Some cells liked horizontal lines, while others responded to vertical lines. An individual cell in the brain is not like a pixel in a digital image. Your brain doesn't see an image that resembles a photograph, but it breaks down images into their components, an amalgam of lines and curves arranged in a variety of orientations. Edges combine to create corners, and from this we get shapes that eventually become faces.
Functions are localized
Each area of the brain carries out a specific function that's crucial to our visual experience. There's a part of our brain called the fusiform gyrus that only turns on when we look at a face. If the fusiform gyrus is damaged, you can no longer recognize your mother by sight. There's also a region that recognizes landscapes called the parahippocampal place area. In one case, a man who had damage to the parahippocampal place area on both sides of his brain was able to get around and recognize faces, but he never knew where he was.
Plasticity is pervasive and crucial
The brain changes throughout our entire life because we're constantly seeing things we've never seen before.
We've only had written language for a small part of human history, so our brains are not genetically programmed to read. When we learn to read, a part of our brain must rewire and become devoted to recognizing words and letters. This adaptability doesn't stop when we're young; you can learn to read at any age. A researcher in China found a group of illiterate Chinese speakers in their forties. Initially, when he scanned them in an fMRI, their brains didn't respond to Chinese characters. After he taught them to read, their brain activity was identical to Chinese speakers who learned to read when they were young.
Sinha described the changes which take place when a child who is born with cataracts has them removed. From birth, "They see the world as if looking through a ping-pong ball that's been cut in half; they are looking through clouds. After surgery, we can see the mistakes that their brain makes as it learns to see." Someone who has seen all their life will recognize the edges of an object. For these patients, a basketball cannot be separated from its shadow initially, but over time they can learn to see edges. "We used to think that the brain was fully formed after 2-3 years of age, but we now know that development occurs throughout life."
The limits of plasticity
At the same time, there are critical periods for development. If the brain doesn't receive exposure during these times it can never learn to respond in the typical way. In terms of visual rescue, the earlier you can start it, the better the outcome will be. If someone's eyes aren't properly aligned at the age of 3 or 4, they will never see the world in three-dimensions; it will always be flat.
Analysis and Future Directions
Artists have been performing tests on the visual system for years. They are doing experiments on our perception. A glance at a Picasso demonstrates just how much you can distort reality but still recognize it. By understanding vision, we'll know what makes these pieces of work so powerful.
We've uncovered some important properties of the visual system, but many more characteristics elude us. Ted Adelson suggests that we won't fully understand the computations that take place in the visual system until we can recreate them. "When we can program a machine to see as well as a human, then we'll understand human sight."
The brain is not static, so it's important to understand how it works at different stages of our life. In addition to mapping out the mature visual system, we need to know how the brain gets there. Sinha believes that when we understand that, we can help patients develop proper vision even if they get a late start in life.
Our senses are important, because they inform our actions. Seeing the world allows us to interact with it.
***Catch the recap of the first episode here.
Epstein R, Kanwisher N. A cortical representation of the local visual environment. Nature. 1998 Apr 9;392(6676):598-601.
Kanwisher N, McDermott J, Chun MM. The fusiform face area: a module in human extrastriate cortex specialized for face perception. J Neurosci. 1997 Jun 1;17(11):4302-11.
Hubel D, Wiesel T. Receptive fields of single neurones in the cat's striate cortex. J Physiol. 1959 Oct;148:574-91.
Tan L, Feng C, Fox P, Gao J. An fMRI study with written Chinese. Neuroreport. 2001 Jan 22;12(1):83-8.