That Neuroscience Guy - The Neuroscience of Mirror Neurons and Learning

Episode Date: February 26, 2023

Humans are great at learning by watching. How do we transform watching something into doing it so well? In today's episode of That Neuroscience Guy, we discuss mirror neurons, cells in your brain that help you mimic other people's behaviour. 

Hi, my name is Olof Kregolsen, and I'm a neuroscientist at the University of Victoria. And in my spare time, I'm that neuroscience guy. Welcome to the podcast. First of all, apologies, we have been away for a week. It's our reading week here at the University of Victoria. And both Matt and I did something a bit unusual and took a week off. So we missed an episode and a bite. As opposed to making up for it, we've decided just to carry on. But as I said last time I was on, the podcast is going to go until just after Easter. Then we're going to take sort of our end of season four break.

Now for today's topic, this is a request from a listener. So again, follow me on Twitter at that Neuro Sci guy and feel free to DM me messages. We're planning season five already, and we're still squeaking in a few things for the end of season four. So one of the requests is for something about what are called mirror neurons. And I've mentioned them once or twice, but we didn't really do a lot on them. And it's part of a deeper question, which is, can we learn simply by watching someone do something? And we know we can. That's sort of a given.

But how does it work in the brain? And that's where mirror neurons come in. And we'll even go back to our old friend dopamine and talk about dopamine briefly. So on today's episode, mirror neurons in the brain. Well, the concept of mirror neurons is pretty straightforward, but it was a surprise to everyone when they found them. We're dealing with research that goes back 30 or 40 years for the preliminary stuff. you know, research that goes back 30 or 40 years for the preliminary stuff.

But it was the early 1990s when they sort of became a real phenomenon. And let's talk about what they are first. Well, first of all, at a high level, what's a mirror neuron? Well, a mirror neuron is basically a neuron that fires when we do something. So it fires when the motor system is doing something, but it also fires when we watch someone else perform an action. So if we see someone reaching for a coffee cup, our mirror neuron system will fire and it will also fire when we reach for the coffee cup themselves. Now they were discovered, like I said, in the early 1990s, although there were some suspicions before

then but the first actual real recordings were from studies in monkey and they were recording from individual neurons within the brain and they weren't actually on the quest to find mirror neurons it was sort of one of these things that they they discovered by accident. But what they found is that when the monkey was watching someone reach for an object or do something or watch another monkey do an action, this mirror neuron system would fire. And so they tested out a bunch of different actions. So, for example, picking up a peanut to hand it to the monkey, some of the system would light up. All right. So the mirror

neurons would light up as well. Now they do respond again when you do the action yourself, but it became very, very clear that they had discovered this mythical mirror neuron system. So neurons that fired when monkeys were watching someone do something. Now, why in monkeys? Well, we haven't talked about all of the neuroscience methods. We did a couple of seasons ago, but one advantage to this type of research, whether you agree with it or not, is you can record from a single neuron at a time or a group of neurons. So you do have to put a wire into the brain. But the advantage to that technique is that you know exactly when the neuron is firing, but you also know exactly where in the brain you are.

With the techniques in humans, things are a lot more vague than that. So anyway, the first research was done in these monkey studies, and they recorded this mirror neuron system, and there's been a lot of follow-up research since then. So in humans, the problem was a little bit trickier because you can't do these single unit recordings, right? You can't ram a wire into a human brain. It's actually not 100% true. Neat, trivial fact. Some people, when they're pre-surgery and they're about to get some form of brain surgery, do volunteer for single unit recordings in humans.

So it does happen. It's just quite rare. But the first work in humans was basically looking at what's called a motor evoked potential. It's basically a way to measure the electrical activity of the brain and determine when the motor system is ready to initiate an action. And what they found is that when people initiated actions themselves, then you saw these motor evoked potentials, and that's been known for a long time. It's just basically your brain.

You're capturing the electrical activity of the brain that's a result of the movement. But what was interesting is they had observed motor evoked potentials when someone was watching the experimenter reach out and grab different objects. So this was kind of the first human evidence of mirror neurons because, you know, here's the motor system firing in response to someone else grabbing an object. Now, since then, this has moved ahead quite a bit. And some of the key examples and the most prominent findings have been using functional magnetic resonance imaging. So we talked about fMRI a while back, but again, just a quick review.

Basically, you put someone in an MRI scanner. You have a special head coil that makes it functional. What you're actually measuring is blood flow in the brain. So you're not measuring neurons firing, but you're measuring blood flow that has been hypothesized due to be a result of neurons firing. And again, I guess you can go see where this is going, but they saw activity when people made simple motor movements. Cause one of the problems, uh, is that you can't do much in an MRI scan. You're lying on your back with this big thing wrapped around your head. But you can do simple motor movements.

And sure enough, when people made movements themselves, you saw the predictable activity in the motor system. But when people saw the experimenter watching figure movements, and this would have been through some sort of viewing mechanism, but again, they still saw neural activity in response to those movements. And that activity specifically was in the frontal cortex and parts associated with the motor system, and also in the parietal lobe.

And if you remember, the parietal lobe is associated with visual for action. We talked a lot about that when we talked about the dorsal stream. And then, of course, the frontal cortex with the motor system made a lot of sense to see those things as well. Now, that's the motor aspect of this. But what's become even more exciting is that people have started hypothesizing about a general mirror system, that maybe this is how our brain is wired to learn. And for example, one of the ways we learn to speak that's been hypothesized is that when we see someone speak, all right, we have a mirror system that fires. And that's one of the ways that we learn this human speech pattern and what's cool

about this is i guess speech is a motor action but the idea is maybe this is just a general thing we do we have these brain neurons that fire and and they capture other people doing things and if you think about infants especially there's been a lot of thoughts about you know developing child where this must be one of the key mechanisms which what they learn like we know they experiment with their motor system but anyone that's been around infants know that they spend a lot of time studying things and this general mirror system could be a key part of learning now i will point out here that support for the general mirror system is a little bit less developed.

With the motor system, it's pretty well assumed to be true. It's pretty easy to capture it with fMRI, and studies continue to go on in monkeys. But another interesting offshoot of this idea of a general mirror system is some researchers have even proposed that this might be an underlying mechanism for autism. People with autism typically have a hard time understanding the actions or the thoughts of others, reading facial expressions and things like this. And if that mirror system is the mechanism by which we understand these things, then having that system at a deficit could explain some of the functional differences we see in people with autism. Now, the mirror neuron system isn't the only way we learn by watching someone do something.

Another key system is the reinforcement learning system that we've talked about in the past. If you think back to what we said, or what I said, I guess guess the way this works is simple anytime we do something our brain sets up an expectation all right so I reach for my coffee cup and I expect to grab it all right or I write a test at university and I expect a grade and then there's an outcome you know I either grab my coffee cup or you know I get I get my test result back. And this computes what we call a prediction error in the learning business. And a prediction error is generally thought to be evaluated in the ventral striatum. And that prediction error then is conveyed to the rest of the brain by the midbrain dopamine system. And then that is used to strengthen neural connections

through a reinforcement learning or in psychology, what you would call an operant conditioning perspective. And this is actually what underlies Pavlovian conditioning in a sense. But from an observational learning perspective, there is human data for this as well. I've got a good colleague who's a researcher in Germany. His name is Christian Bellbaum, and he's published some cool results using human EEG, so brainwaves, where he shows that when someone does something or performs a task, if you will, usually these are little games you play on the computer, when they get it wrong or the outcome isn't what they expected, you see a difference in the EEG

signal that is associated with this release of dopamine. In my own lab, this is one of the big things we study as well. This has been seen for a while, but what Christian did that was really cool is he saw that when people watched someone performing these cognitive tasks, he saw very similar brain activities. So he saw evoked brain responses to someone else doing something. Now it's important to note here, this isn't the mirror neuron system. This is a well-studied phenomenon that's associated with this computation of prediction errors that are evaluated by the ventral striatum and other parts of the brain and used by the anterior cingulate cortex and other

parts of the brain to learn. So two ways that we can learn through watching people do things. One is the mirror neuron system. When we watch people do actions, the system fires. And the hypothesis is this is one of the ways the motor system gets information about, you know, how do I do this if I do this myself? And then the other is this prediction error system. So the reinforcement learning system we talked about way back in season one, where when someone else does something, you set up an expectation of what's going to happen. You see an outcome and you get a prediction error. Now, one thing I will emphasize with this, because this comes up when I teach this at the

university, you can't watch yourself to expertise. All right. So these systems do capture these actions, but you can't simply learn by doing this. So if your thought after listening to this was to turn on a basketball game or go out and watch someone play tennis and you would learn and become an expert, that's not going to happen. Your motor system is going to get the general idea of the movement, but of course you still need to practice and rehearse it yourself. So there's a whole bit on how we learn from watching things, answering the requested question on mirror neurons. I hope that helps. I hope that was what you wanted to know. At least there's enough information there. Also this dopaminergic system playing a role. So that's what I've got about how we learn from watching people do things.

Thank you so much for listening as ever. Of course, check out the website, thatneuroscienceguy.com, the Etsy store, T-shirts, all the money from the T-shirt sales goes to my graduate students in the Kregolson Lab. There's also Patreon. Thank you so much to those of you that subscribe to Patreon,

a dollar a week, a dollar a month, any pledge, it comes off your credit card automatically. But again, we pull the money out and all of it goes to supporting graduate students doing neuroscience research. There's Twitter. I mentioned that at the outset at that neuroscience guy,

a great at that neuroscience guy, sorry, a great way to send us show ideas or just to communicate and finally thank you so much for listening to the podcast we're closing in on 350 000 downloads we never would have guessed that two years ago because we're about the two-year anniversary of our first episode so thank you so so much for listening it inspires us to do another season and keep going please subscribe of course. My name is Olov Kregolson and I'm that neuroscience guy. I'll see you on Wednesday for

another neuroscience bite.