Our last example of a connection between a brain map and perception, concerns visual motion. And I'm going to tell you about an experiment I conducted in collaboration with my friend Rick Born. When we were both post doctoral fellows working with Bill Newsome in the 1990s. We were interested in how these brain maps could be used to compute things and to guide behavior. Could we unpack the connection between neural activity, on the one hand. And perception on the other. To tell you about the experiment we did, I need to tell you a little bit about our sensitivity to visual motion. The brain area that seems most likely to contribute to this sensitivity. And the behavioral skill that we all possess, and that relies on sensitivity to motion. I'll be putting my eye movement hat on and telling you about how our eyes move in response to moving objects in our visual scene. So we're very sensitive to moving stimuli. Things that move are more likely to be alive. And that may mean they are either tasty to eat. Or pose some kind of threat to us, like lions and tigers. Or simply cars on the highway. So, our brains have evolved an ability to pick up on motion and emphasize it. But this perceptual quality of motion is something that really comes from within our own brains. At the risk of belaboring something that might seem obvious, a moving stimulus is something that changes in position over time. But there are many things that change in position over time that we don't actually perceive as moving. Things that change in position either too slowly or too quickly for us to process in this way. The hour or even the minute hand on the clock for example, move too slowly to be classed as motion. Here's a short time lapse video of something that certainly involves changes in the visual scene across time. But which we don't normally perceive as motion. But here, sped up, you'll be able to see the shadows of tree branches moving across the snow. [MUSIC]. And I think it's interesting in that video, that the shadows of the tree branches seem to be moving. But stimuli that we normally see as moving, like people and cars. You just see as isolated, momentary snapshots. Stimuli can also move too quickly to be perceived as moving. And here's a slow motion video of colored paint poured on top of a stereo speaker. Normally, the membrane of the audio speaker moves too quickly to be perceived to the naked eye. But here, slowed down, you should see it very clearly. [MUSIC]. Now the brain area that seems most likely to be responsible for our sense of motion, is a cortical area called MT. It's given the name MT, because it's located in the middle temporal region of cortex. And this line shows where MT is located in comparison to primary visual cortex. Where V1 and several other visual cortical areas. >> So primary visual cortex is located over here. And area MT is located in this general area. In human patients, lesions of area MT, cause impairments in perceiving moving stimuli as in fact moving. Studies of the neurons in MT have revealed three important features. >> Overall, the neurons have receptive fields in space and form a retinotopic map. But neurons are more responsive to moving stimuli than to stationary stimuli. And they are sensitive to the direction and speed of motion of those moving stimuli. Furthermore, they're topographically organized according to those velocity preferences. So there's a kind of a velocity submap located within an overall retinotopic map. So for every location in the visual scene, there are neurons and MT, that are particularly sensitive to moving stimuli at that location. But sensitive to the direction that a stimulus might be moving. Here's a demonstration of this. This video is a recording of an empty neuron's responses to moving visual stimuli. And you can watch and listen. And try to figure out which direction you think this neuron responds best to. [SOUND]. The last piece of set up to our experiment that I want to tell you about, is about a behavior that we all engage in when we see something moving. And in fact, we can only engage in this behavior when we see something moving. It's called smooth pursuit eye movements. And the goal of smooth pursuit eye movements, is to help us keep moving targets stable on the retina. So that we can see them more clearly. So I'm going to try to demonstrate, smooth pursuit eye movements to you. But before I do that I need to put on my special eye movement hat. So you're going to get to see how my eyes move in response to a moving target. These are all little eyeballs. But, what you're going to see is how these eyes move. Not just these eyes. Okay. So on the next slide, there's going to be a moving target. And I'm going to try to track that moving target. And to show you my eyes more clearly, I'm going to take the camera. Bring it around. And hopefully you'll be able to see my eye movements. Alright, that looks reasonably well lined up. Reasonably well lined up. Okay. And, so here we go. [BLANK_AUDIO]. Let's see, I'm going to be looking down here. [BLANK_AUDIO]. Alright, well hopefully, that came out. And the cool thing about smooth pursuit eye movements. The cool thing about smooth pursuit eye movements, is that you can't actually make smooth pursuit eye movements unless you see a moving target. If I stop this target from moving. If I stop that target from moving,. And I tried to make the same pattern of eye movements just from memory, I wont be able to do it. My eyes will actually jump from place to place, making those sicatic eye movements that we talked about in an earlier lecture. So smooth pursuit eye movements are a really good way to assess what the brain is perceiving. And how it is constructed from those signals in areas such as area MT. Okay. So now you know that we perceive motion. That there's an area of the brain that is uniquely suited to encoding that motion. And that there's a particular behavioral response that mimics the motion that we see. So now I'm ready to tell you about the experiment that we did. So one puzzle is that there are many, many neurons in the MT. Each one of those neurons can be thought of as trying to signify that there's a particular direction of motion happening in the world. We'll call that its prefered velocity. What we didn't know, was how that population of neurons computes a single answer. That can be used to guide the eyes, to move smoothly at a particular velocity. It's kind of like there's an election and there's many candidates running. And every candidate, namely every velocity vector, is getting some votes from some neurons. And the question is, how do you tell when you have those votes? Well, there are three possibilities that we considered. One is that you just simply look at what velocity vector is getting votes from the most neurons in MT. That's called the winner-take-all algorithm. A second possibility is that the brain might consider all the velocity vectors that are getting some votes, and compute the average of those velocities. And the final possibility is related to that average. And it is, take all those velocity vectors. But instead of computing the average, compute the sum. Any of these algorithms could be calibrated to work properly when there's a single moving stimulant. And all the brain has to do is program an eye movement to track that one visual stimulus. And that's the situation that normally arises. Because there can only be one visual stimulus at a particular location in space. But we realized that we could fool the brain into trying to encode more than one velocity, by using electrical stimulation to cause neurons and MT to be active when they otherwise wouldn't be. So we did an experiment in which we presented a moving visual stimulus. And we also electrically stimulated an area, MT. To introduce an electrically induced illusion of motion. Then we could look at what behavior resulted. In winner-take-all, is the method that the brain uses to program an eye movement based on the signals in MT. Then what you'd expect to see is that there might be some trials in which the subjects would pursue, just as if the only thing that they saw was the real moving visual stimulus. But there would other trials in which they'd pursue as if the only motion signal was the electrically induced artificial one. Under the averaging hypothesis, what you'd expect to see is that the smooth pursuit eye movements. Would reflect the average of the real visual stimulus that was moving. And the artificial electrically induced motion signal. And under summation, what you'd expect is that the smooth pursuit eye movements should reflect the vector sum of those two motion signals. To make a long story short, what we found was most consistent with the averaging hypothesis. Subjects generally tended to pursue a motion vector that was intermediate between the true visual stimulus, and that artificial electrically induced motion signal. It's as if the brain is trying to arrive at a consensus, by looking at all the motion signals that are present in area MT. Now one thing we still don't really know, is how the brain actually computes that average. Computing an average is actually a rather tricky thing to do. It involves dividing one number by another. Although the arithmetic of that is easy. It's not clear how neural signals accomplish that using synapses like we've talked about in previous videos. If you're interested in learning more about that problem, here's a paper that I wrote about a few years ago. That has some computational models, proposing ways that the brain might actually perform this calculation. Okay. So one thing that we get asked fairly often is, do we think that the subjects perceived motion in response to the electrical stimulation? We don't know for sure because the experiment that we did was conducted in animals. So we couldn't exactly ask them. But I kind of think so. And the reason, it comes back to those smooth pursuit eye movements. Since you can't make smooth pursuit eye movements unless you see a moving visual stimulus. It seems likely that they experienced a perception of visual motion triggered by the electrical stimulation. So now you know a good bit about vision, touch, body position sensing and maps in the brain. But there's one spacial sense tha we haven't talked about at all yet. And that is the auditory system. The auditory system is really quite different from vision and touch. And you'll see that the brain has to work pretty hard to accomplish some of the things with sound, that it does so easily for visual and tactile stimulae. So we'll turn to the topic of sound in the next video.
