PAUL KOCH: I'm an assistant professor here in the Department of Neurosurgery.

Deep brain stimulation is a technique and a technology.

It is a technology where we can permanently implant electrodes into very specific structures in the brain, and those electrodes are then connected to a smart device basically, a tiny computer and battery that gets implanted under the skin and controls the kind of electrical stimulation that those wires can deliver.

We now have a number of different diseases that we treat with deep brain stimulation.

The most robust and the one that's been around the longest is movement disorders, such as Parkinson's disease, and essential tremor, which is another form of shaking or tremor that people can develop.

And when medications that have been developed to treat those diseases no longer are doing a great job at controlling their symptoms, they often are candidates for implantation of a deep brain stimulator, or we usually use the term DBS for short.

These have had great success in the movement disorders world for decades.

They can take someone who has become debilitated by shaking, by rigidity and stiffness, being unable to eat or do a lot of the activities that they enjoy, to returning to those activities for many, many years.

So more than a life prolonging therapy, this is really a quality of life improving therapy and it's been very successful on that front.

More recently, we've started to use the DBS technology for epilepsy, and epilepsy is a condition where a patient will have seizures spontaneously that don't seem to be provoked by anything in particular.

So like for example, if somebody has a seizure after a car accident and they hit their head or they develop a bleed inside the brain and they might have a seizure, those are considered provoked seizures.

But when a person develops a condition where those seizures just happen spontaneously as they're going along through their life, we call that epilepsy.

And again, about 20 to 30% of patients who have that condition are poorly controlled with medicines.

And so again, we can use DBS, as well as some other technologies, to attempt to arrest seizures either as they're happening or before they happen.

And we've been enjoying very good success with these technologies in returning quality of life to patients.

So in the case of epilepsy, not only are we improving quality of life because they're not having seizures that can impair their ability to lead independent lives, hold jobs, live on their own, drive, but we also know that if your seizures are not controlled, you're at higher risk for injuring yourself and also for death.

And so these technologies are both improving quality of life and they're improving the risk of injury and mortality.

So they've been a real game changer for epilepsies that we can't treat other ways.

My research has to do with how does a brain go on to develop epilepsy over a period of time after some sort of trigger.

So some epilepsies are caused by a head injury, and we call that post-traumatic epilepsy.

And these people, they will suffer a brain injury, and they may go for a couple months or even a year or two without having seizures and then they will suddenly start to develop them.

And so the question in my laboratory is, one of the questions in my laboratory is, how does a brain go from an initial single insult or brain injury, how does it evolve in time, months to years later, to then all of a sudden start having seizures?

We don't really understand that process.

Despite knowing that it happens for a long time, we don't understand how it occurs, who's going to develop it.

We have some risk factors that have been identified, but nothing that's a slam dunk.

And we don't have any therapies to prevent it from happening.

So the ultimate goal of the work is, one, can we understand how the brain networks, how the different parts of the brain are communicating and how that communication fails or goes awry to ultimately lead to epilepsy.

And then number two, can we identify nodes in those failing networks to intervene upon to prevent epilepsy from happening altogether.

Because all the therapies we have now for people who develop epilepsy are after it's developed, it's to try to stop seizures from happening or to reduce the number that they have or the severity that they have.

But the goal of the laboratory is really to try to prevent it from ever occurring.

We talked a little bit about deep brain stimulation.

This is one device in a collection of devices that we might refer to as neuromodulation.

And that's sort of the goal.

The goal is to modulate the neural activity in the brain in a way that's helpful.

DBS is a relatively dumb system, meaning I can set some parameters and then turn it on and it will stimulate in a pattern that I've chosen at a frequency that I've chosen.

And we sort of titrate that based on experience and how people do as to what the optimal settings are to achieve the goals that you're going to achieve, whether that's stopping somebody from shaking or stopping seizures from occurring.

We have a more sophisticated device that we use in epilepsy called RNS, or responsive neurostimulation.

This is a device that also involves implanting electrodes into the brain, but we first have to figure out where we think the seizures are starting in the brain.

So, many types of seizures start in one place in the brain and then spread along a larger network.

And oftentimes, one of the goals of treating epilepsy from a surgical standpoint is to figure out where those seizures start and to either remove that part of the brain because we know that that actually gives patients the best chance of being seizure-free.

But if we can't do that either because it's too large an area or it's in a very important area is that we can't actually remove it without causing a lot of problems for a patient, then we sometimes go towards this device called responsive neurostimulation, where we implant an electrode in that spot or that area where we think the seizures may be coming from.

And the device is then able to listen to the brain.

So it can actually, it can to a limited degree detect when a seizure's beginning.

And when it does that, we program it to say, okay, let me zap it, essentially deliver a little electricity to shut down that seizure from spreading or from continuing.

And it's been a very, very successful device.

For patients that we can't cure by removing the part of the brain that's causing the seizures, this is a great alternative and can reduce the seizure burden that patients have even up to 60 or 70% over the long term.

So it's been a great device, but it is a what is called a closed loop system, meaning it is autonomous in the sense that it's listening on its own, it's detecting a seizure by itself, and then it's acting on that detection by delivering stimulation.

But it's a relatively dumb closed loop system, meaning that we have to tell it what it's looking for.

We have to program it and say, you know, you're looking for the brain activity to change in this way.

And when you see it change this way, then deliver stimulation.

I think the next generation of these kinds of devices, which I think will be dependent on AI, because what is AI?

It's essentially very sophisticated pattern recognizers, right?

That's what ChatGPT is.

It's using this vast amount of information that's been fed into it, and it's learned patterns and it's learned how to say, oh, when you say, what do you want from the store, or I'm supposed to say, oh, I want some food, or something like that.

They're just sophisticated pattern recognizers that are very powerful.

And so the next generation of these devices I think are going to be able to figure out on their own what is that pattern in the brain that I need to act on?

And we are not there yet right now, but I think we will be incorporating AI into these kinds of devices going forward, for sure.

I mean, it's all a continuum, right?

It goes back many, many decades when the first electroencephalogram were figured out, and that's a term that refers to measuring the activity of the brain, measuring the electrical activity of the brain.

So that came about at a time when technology was at the point where we had sensors that were sensitive and sophisticated enough to detect the slight changes in electrical activity in the brain.

And that's decades and decades old.

So it began there.

And then I think that, you know, the big step for something like DBS or RNS was when computing power, storage, batteries, the ability to have a computer chip process information got to the point where it could be miniaturized, small enough that it could be implanted in the body, I think that drove a lot of this.

But the ideas behind them, the idea of being able to stimulate the brain to affect a therapeutic change has been around for a long time.

People were doing this in the hospital, so long, long ago, to try to treat many, many different diseases.

And there is a long and somewhat sorted history of some of this, of putting electrodes into parts of the brain for patients suffering from many different types of diseases and trying to improve their symptoms.

So that's been done for a long time.

But when we were finally able to do it on a permanent basis and have a little device that someone can walk around and go about their normal life, that I think was a key moment in which these things really started to benefit a lot of people.