Home - Brain - Ethics and the Neural Code
Feature

Ethics and the Neural Code

Brain mapping combined with electrical stimulation presents both promise and peril, which the neuroscience community is taking seriously, according to one pioneer

By
Jason Socrates Bardi, Editor-in-Chief
Fri, 05/26/2017

B2D-yuste2.jpg

Yuste is wearing a red shirt and sitting in the grand lobby of the Moscone West convention center at dusk

Rafael Yuste, a neuroscience professor and the director of the Neuro Technology Center at Columbia University sitting for an interview on January 29, 2017 at the SPIE Photonics West meeting in San Francisco.

Jason Bardi/AIP

SHARE: 

Modern medicine has yet to decipher the neural code—those rules of the game that relate the spatial and temporal patterns of our neuronal activity to our higher behaviors and mental states. But the most recent results of high-resolution, 3-D brain mapping animal studies have shown great promise. We can read distinct patterns of activity, disrupt them or even write new patterns with specific meaning and content.

To many, it's a wonderful new world. Emerging neuroimaging technologies promise to impact research, medicine and society and revolutionize everything from the treatment of mental diseases to our philosophical understanding of consciousness.

To some, it's also a brave new world. Those same recent results portend the peril that could come through the improper, unethical use of such tools. That suggests a clear need for ethical guidelines to calm fears and ensure technologies developed in the coming years will truly be for the benefit of humanity, according to Rafael Yuste, a neuroscience professor and the director of the Neuro Technology Center at Columbia University.

We sat down with Yuste at the SPIE Photonics West conference in San Francisco on January 29, 2017 to interview him for an upcoming video for AIP's Inside Science News Service, which is owned and operated out of AIP. This is a partial transcript of that interview.

How do you develop ethical guidelines for future technologies when you don't know what the resulting technology or its uses—or misuses—could be?

We actually organized a workshop at Columbia in September [2016], where we invited top experts in neuro technology today with experts in ethics and bioethics, exactly to address these issues you raise: What is likely to happen? What things are in the pipeline? And the overwhelming sensation of all of us who organized, who were at this workshop is that even today, there’s already methods out there that you have to be watchful for.

And the things that are in the pipeline are things that are very likely going to have a huge impact on society. So it’s not the science fiction realm anymore, and I think it behooves us as not just scientists but responsible citizens to think ahead of the population and alert society of what is coming so that it can be properly integrated into the world.

You mentioned technologies that are in the pipeline. Can you talk about some of the technologies that are coming down the road —and maybe even some that are imagined.

I’m going to focus, for example, on the possibility of reading people’s minds. So today there’s a group here, actually at Berkeley across the [San Francisco] Bay, that is using fMRI… They can take a person, put them in a fMRI machine, show the person a bunch of images, a hundred different images, and then scan the brain, and predict what the person is thinking of. Okay.

Essentially there’s nothing that medicine can do to fix those problems because we don’t understand how these diseases work

And this is done with a method, fMRI, which has very poor resolution—both temporal and spatial. In this meeting, the work that we and other people are doing with mice enables us to, with sophisticated lasers, read out the activity of hundreds or thousands of neurons in particular parts of the brain.

At some point these methods are going to be developed enough so they could be applied to humans.

So if we can already decode the images in the minds of people with fMRI, imagine what we could do if we could actually completely read the activity of brain tissue. So this is in a way something that needs to happen, because of the patients. You know, we have a lot of patients in the world that suffer from mental disease and neurological diseases, and essentially there’s nothing that medicine can do to fix those problems because we don’t understand how these diseases work.

The methods are critical to do that. We have to use these methods, [and] apply them to human patients. That will help us understand their diseases and help us cure them. But the same exact methods could be used, for example, to read out the minds of people. And we have to make clear that just like it happens in medicine where there’s some methods that could be used to harm people, but they never have, because ever medical student recites the Hippocratic Oath throughout the world, generation after generation. The same thing should happen with this technology. And there should be some sort of oath or, or deontology that people are aware of and they respect and is spread out widely.

So you talked about reading thoughts. What about memories? Can you help people suppress bad memories?

Let me give you another example of what’s in the pipeline. So, again, in mice today there are groups that are implanting false memories into mice using a very clever sophisticated genetic method. In our own group at Columbia, we’re using optics, with optogenetics, to implant patterns of activity into the brains of mice and these patterns are stable after days…

If you can go in and alter the firing of the neurons of the brain, you may be able to alter the concept of self, the concept of identity

You can imagine that this method on the one hand could be really useful to treat for instance schizophrenics, if there’s a way to put in a pattern that cures the abnormal patterns that they have. It would be a little bit like rechanneling a river that has overflown its banks. But the same methods could be used, if we decipher the meaning of the neural activity—if we break the neural code—the same methods could be used to implant memories and ideas, just like we’re doing, as a field, in mice today. So that’s another example of what’s in the pipeline and that’s something we have to think through carefully now.

What the implications are…  

Exactly. Actually very profound issues come up. For example, who controls the access of this technology? Can you use that to enhance the performance of humans? Who should regulate this? Should governments allow individuals to decide whether or not they want this technology applied to themselves? Or is it something that the medical profession or a future type of expert should [decide]?

Other issues, which are even deeper, have to do with the concept of self. Obviously, we think of ourselves as being one. Our identity is critical to who we are, and this, at the end of the day, is built by the firing of the neurons in our brain. If you can go in and alter the firing of the neurons of the brain, you may be able to alter the concept of self, the concept of identity.

I want to ask you about decoding the brain. First of all, would you define for me, what is the neural code?

The neural code is, at the end of the day, something very simple. Neurons are talking to each other using a language that we don’t understand. It’s just like understanding a foreign language—we have to break in, have a dictionary that lets us know what’s the meaning of the words. This is exactly the same thing that happened in genetics. In genetics, in the 1950s, we knew that DNA had the genetic code, we knew that it was written with nucleotides, but we didn’t know how the nucleotides and groups of nucleotides corresponded to the structure of proteins.

So through a set of very clever experiments in England, Crick, Brenner and Barnett figured out that the genetic code had three letters. And after that it was all downhill—they cracked the code, and this is the birth of molecular biology.

You look at the heart, the lungs, and a few centuries ago we had no clue how they worked, we said, "oh my God, it’s too complicated!"—well guess what, we did it

We cannot, in any animal, we cannot read the neural activity and predict what the animal is doing, or what the animal has done, or what the animal wants to do. But it will be broken. I have no doubt.

This is just another biological challenge that is going to be conquered by science. I hope to be able to see that. It’s going to be very exciting, just like it was in genetics when they broke the genetic code.

If you consider there are tens of billion neurons in the brain, some of which make tens of thousands of synaptic connections to other neurons, is the brain just too complicated?

I don’t think that understanding the neural code and cracking what the brain does is too complicated. And why am I saying that? Well, for two reasons: first, look at the success of science and medicine over history. Little by little we’ve conquered and understood every part of the body… You look at the heart, the lungs, and a few centuries ago we had no clue how they worked, we said, "oh my God, it’s too complicated!"—well guess what, we did it. The brain is just another organ. There's no magic, OK. It's not like we have something there that is extraterrestrial.

People often talk about precision medicine for its benefit to the patient. Do you envision decoding the brain will enable more precision medicine approaches?

Yes. Let me give one example. So in our lab we’re studying mice that are models of human schizophrenia. So they have genetic defects, like some of the human schizophrenics have, and we can also generate a pharmacological model of schizophrenia in mice. Now I’m not saying that these mice are schizophrenic, but they resemble, in many ways, the patients.

So we’re mapping the activity of the brain of these mice. And we’ve discovered that the neurons look normal, they fire normally, but when you look at the way in which they fire together, there’s something very distinctly abnormal.

You find out what the problem is, and then you go in with other methods to correct that problem

Let’s imagine for a second that this is what goes wrong in the brains of schizophrenics—that the way the neurons fire together is the problem. So if this is the case, then I could imagine a precision neurology or precision psychiatry in the future where you use some of these methods to map the neural circuits—the activity of neural circuits in schizophrenics—in individual patients. You find out what the problem is, and then you go in with other methods to correct that problem.

Let’s say that there are some neurons that are firing abnormally, maybe you can silence them. Or the neurons are not firing enough, well maybe you can actually stimulate them to fire. And it’s a little bit like correcting the flow of neural activity, rechanneling it into a normal pattern of activity which should relate into a normal pattern of thought.

Now I want to be very clear we don’t have those tools yet, we’re not there yet, and this is a hypothesis. So I’m talking about basic science before we can actually go and really help the patient.

Nevertheless, it's amazing to think about those things. So my next questions is: what do you think is around the corner in the next 5-10 years?

In terms of the methods, I think we will have methods, probably optical methods and methods coming from nano-electronics that will enable us to record the activity of millions of neurons in awake behaving animals and maybe in human patients. I think that’s going to happen.

At the end of the day I’m pretty sure they’re going to say, you know what, it was actually quite simple.

And these methods will stimulate a new generation of computational scientists and theorists to try to crack the neural code. So I’m, I’m bullish. 

I think the neural code could be cracked in the next five to ten years. We will wake up one day with the excitement of opening up a journal and reading how someone figured this out—just like it happened in the 1950s with the genetic code.

At the end of the day I’m pretty sure they’re going to say, you know what, it was actually quite simple. Just like it happened in genetics, where people were going crazy with all kinds of weird hypotheses, and it turned out to be something very straightforward: just a simple double helix and that was it, no?.

So, so I’m living for that moment. I hope to see this and I hope it happens in the next five to ten years, that the new tools will enable us to crack the code, as a field, and this will turn a corner in mankind. We’ll be able to look at our minds from the inside scientifically for the first time, understand who we really are.