Four impossible things before breakfast: Amy Bastian’s empathic logic

I had wanted to write up Amy Bastian‘s excellent 8:30 am talk from SfN, and finally it is happening. That’s right, my hangover and I were front and center so you didn’t have to be. I’m looking at you, #sfnbanter. You’re welcome. A couple of weeks late.

This talk was excellent not least of all because Bastian understood that at 8:30 am, no one was functioning quite properly yet. There was going to have to be a lot of hand-holding, and a lot of balloon walkers. Bastian’s talks are always clear and engaging, and this one was particularly merciful to boot. Bastian walked us gently through four of her papers (instead of a thousand like some other speakers who shall remain nameless), and that is why I retained enough to write you all this (let’s be honest) listicle.

I will start with a point she finished with, which is that many injuries like stroke, trauma, or even surgeries that remove up to half of the brain to stop epilepsy, cannot be treated with drugs, operations, or even fancy stuff like lasers or tampering with DNA. All of which is to say that, and I’m paraphrasing Bastian here, we must engage in some rational way with the patients’ motor processes. We must play their game to figure out how they work, get them to give up their secrets. If you think you’ve heard me gush about this before, it’s because you have. I am a sucker for a paper that reads like a Roadrunner vs. Coyote storyline. That, and I once had a total meltdown upon breaking a piece of equipment that was worth more than I made in a year, so no, I don’t like using fancy equipment.

But physical therapists like Bastian, in particular, wow me with this incredible empathy. Bastian says, it’s easy for friends and family to shout at their injured loved one, “you can do it!” or “just take a bigger step!” Indeed, this is how we encourage babies. But people aren’t babies–their brains are not the doughy, plastic things that babies’ are. They are a loaf of bread with a jagged slice taken out of it, and you need to make a learn-to-walk-again sandwich.

So in this case–and this is what I love about this research–the rational, logical thing to do is to accept what is and work with it. You can’t make someone heal by telling them what to do, and there are no guarantees they’ll heal fully or only partly. This letting go of control is probably the hardest thing asked of scientists, Type A monsters that they are. To promote healing, often it is the brain’s implicit, rather than explicit, learning mechanisms that must take the wheel. You don’t learn to do a cartwheel by poring over an instruction manual, and you don’t learn to walk again by having your mom shout “Walk, baby, walk!” at you as you struggle to rise, unfortunately. You have to play by the rules of whatever brain part is still working, and these might be radically different from what you are used to, as a person who used to have a much wider repertoire of movement capabilities.

Bastian’s work is focused on figuring out what those rules are and playing by them. During her talk, I saw a little girl with half a brain walk around like any other kid, like it was no big deal. She should be paralyzed on one side of her body and who knows what else. But if you’re wily and you listen, you might think up the right question to ask the motor system, and it might say, ok, you got me. Here are my secrets. She can walk if you do x, y, and z, and then bam: walking girl.

Here are some crazy secrets of the motor system I learned very early one morning a couple of weeks ago.

1. Think of your brain as your 16-year-old self, learning to drive a car, except instead of a car it’s your body. At 16, you figured out how to drive a car, and as a baby, your brain figured out how to drive your body. The cerebellum creates a match between your brain and your body, even when your body changes. Depending what parts of the cerebellum are damaged, you might think parts of your body have more or less inertia, and therefore think they’re heavier or lighter than they really are. Bastian’s team modeled each patient’s “too light/too heavy” body/brain mismatch and corrected for it using robots that hold the arm as it moves. With their arms guided by the robots, the patients’ biases were corrected and their movements became more normal. These models could also be run in reverse: they could program them into the robot arms, strap in a healthy person, and get the healthy person’s arm to move as though they had a bias like that of the cerebellar damage patients. This is promising–in the future, perhaps prosthetic devices could capitalize on these models to provide corrective force to patients’ movements.

[Bhanpuri, Okamura, & Bastian 2014]

2. We rely on predictive control–that is, the ability to predict the consequences of our actions. When we make a mistake, we use that information to update our predictions. But patients with cerebellar damage can’t. This could be for two reasons: It’s possible that the cerebellum generates quick-fix adjustments to movements as we make them. If this is so, patients’ poor ability to adjust is because they can’t come up with these motor adjustments. It’s also possible that it deals with the senses more than movement, generating predictions about what sensory information to expect. If this is so, patients lack the proper predictive power to select the right motor plan for the sensory feedback they’re trying to get (say, shooting a ball into a hoop or bringing the coffee mug to their mouth). Well, it turns out we update our sensory predictions. They figured this out by having patients and controls make “shooting” and “pointing” movements. In shooting movements, they had to point straight to a target while their feedback was manipulated, eliminating the possibility of correcting for the manipulation on the way to the target. In pointing, they were allowed to correct on the way. Patients were impaired at both, meaning that it can’t be the motor adjustments driving learning–it made no difference when they were allowed to make them. Instead, if our sensory system learns that no, our arm will go over there,  not over there, when we engage certain muscles, then the next time we have that goal we’ll be able to pull up the right motor trajectories for that goal. So learning to predict the future, the future within our arm’s reach anyway, is to learn to control it.

[Tseng, Diedrichsen, Krakauer, Shadmehr, & Bastian 2007]

3. On to some of Bastian’s split-belt treadmill stuff. She asked, why do we learn? Why don’t we just keep limping after an injury–what’s wrong with that? Well, limbs are expensive, metabolically speaking. If you walk symmetrically, you exert much less effort. Bastian’s team measured people’s carbon dioxide output as they walked on a treadmill that was split down the middle, causing their two legs to walk at different speeds. Remarkably, people can deal with this and even get so good at it that their walk goes from a limp to a swagger. You hardly notice their legs have different step sizes, the walk becomes so smooth. Crazily, people can also walk forward with one leg and backward with the other. They found that when you learn a new speed, it is learned in a leg-specific and a direction-specific way. What you learn walking forward, your motor system doesn’t automatically assume will apply to walking backward. Ditto for the right and left legs–they each seem to have independent systems for learning. With this specificity of learning, the brain does upkeep for its expensive locomotion habit by straightening things out whereever needed, should we become injured, wear a pair of high-heeled shoes, or gain or lose weight. This knowledge can be leveraged to help customize people’s rehabilitation regimens, targeting the limbs and movements that need it most.

[Choi & Bastian 2007]

4. Now, this last one is my favorite, so, congratulations to you for reading this far. Imagine your friend is injured. What do you do when you go see them in the hospital? Do you offer to help them out of bed, or do you poke at them as they try to get up and laugh? I hope you said you help them, you monster. Well you’re wrong! I mean, no, you’re right, but not as right as you’d think. To rehabilitate someone, you’d think it would be a big help to hold their hand and walk them through whatever it is they’re relearning. That’s how you learned to ride a bike, right? Someone running along behind you, holding the seat? While this is nice, unless you intend to follow your friend around forever, you better knock it off. Patients learn better if you exacerbate their errors. That’s right. If someone’s limping a bit one way, push them a bit farther that way (Clarification: don’t you push them, and don’t tell anyone I told you to–this is what their PT should be doing, you dingus). This will help them learn to push back. If they’re making do with a limp, push them off-kilter enough that they have to catch themselves. Importantly, though, be gentle–the error you cause has to be within a somewhat normal range for them. If you push someone so far they’ve never experienced pushing back that much, they won’t be able to do it. You’ve got to push them to adjust more frequently for errors that are above-average in size. Eventually, they’ll be able to correct for their limp.

[Torres-Oviedo & Bastian 2012]

And so, dear readers, this is of course the part in my rhapsodizing where I tell you What It All Means, because this pushing a limping person thing, it is a metaphor for life. When life knocks you down, all you want to do is hole up in your comfort zone with some Ben & Jerry’s and Toddlers & Tiaras National Geographic documentaries. But no, you need to get out there, slugger, and push yourself, but only within your limits. Try new and difficult things, but be realistic about what you can do. It is the only way you will heal.



This entry was posted in Science. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s