Touching the Void - Part 1

By Dr. Lise Johnson (CSNE Education Manager)

brain and spinal cord diagramWhen we think about spinal cord injury (SCI) we often, quite naturally, focus on the fact that it causes paralysis. One of the most obvious effects of an SCI is an impaired ability to move. Consequently, when we think about improving the quality of life for people with SCI, we think about ways to restore that ability. In the context of neural engineering this often means building motor neuroprostheses – devices that capture brain signals and use them to control an external device (like a robotic arm). This is, of course, very important, but it turns out to be only half of the story. Damage to the spinal cord doesn’t only affect the brain’s ability to send messages to the body; it also affects the body’s ability to send messages to the brain. It’s not just that you can’t move; you also can’t feel. Specifically, SCI impairs somatosensation – a collective term for the senses of touch, temperature, pain and body position. Information associated with these senses is collected at the periphery and makes its way to the brain via the spinal cord. When the spinal cord is severed, none of that information gets through. If you are trying to make a brain-controlled motor prosthesis for someone with an SCI you really need to keep this in mind. Why? Because those senses aren’t just a nice way to develop a rich, full, experience of the world – they’re also very important for movement control. In fact, movement control absolutely requires some sort of sensory feedback. If you have ever gone to the dentist and received a Novocain shot you may know this already. When your tongue and cheeks are numb (which is what you want, if you’re having a cavity drilled), it’s difficult to talk, or eat, or keep drool from running out of your mouth and down your chin. Therefore, if your goal is to build a successful motor prosthesis you need to incorporate some sensory feedback.

Now, I said you need feedback, but it doesn’t necessarily have to be somatosensory feedback. Just because nature provides one solution to a problem that doesn’t mean it is the only possible solution, although usually it’s the best. In fact, most existing motor prostheses have no somatosensory components. Instead they rely on a workaround – they use visual feedback as a substitute for somatosensory feedback. This works fine for humans because vision is our dominant sense, but it is far from ideal. First of all, you have to look at your arm whenever you want to use it. Try doing this for say, half and hour, and you’ll see how irritating that is. Furthermore, there are some situations in which what you can see just isn’t good enough - like when you want to pick up an egg. If you can’t feel how hard you’re pressing on the shell, you won’t know how hard you need to grip. If you grip too hard you’ll crush it, if you don’t grip it hard enough, you’ll drop it. Either way, you have a broken egg and probably a mess. Clearly, it would be nice to have some feeling in the fingertips of your prosthetic hand. Is it possible to get information about the force generated at the fingertips back to the brain? It turns out that it is, but in order to tell you about it I need to start somewhere unexpected; I need to start by talking about epilepsy.

The term ‘epilepsy’ does not describe a disease so much as it describes a symptom. Having a diagnosis of epilepsy just means that you have seizures. Specifically, it means that you are somewhere in the middle of a continuing series of seizures. It doesn’t say anything about the underlying cause of those seizures, what kind of seizures they are, how frequently they occur, or how severe they are. It’s really an umbrella term that describes a diverse group of underlying physiological problems (or pathologies) that manifest themselves with an even more diverse set of physical symptoms. For example, people are born with some forms of epilepsy while others develop epilepsy after a traumatic brain injury, brain tumor, disease or fever. Some seizures affect a small part of the brain while other seizures involve the whole brain. Seizures may cause violent convulsions (what most people think of when they think of a seizure) or they may cause the person to lose consciousness, or space out for a short period of time. The seizures may cause the person to experience strange smells or have visions or feel more connected to the universe. Some seizures are so extreme that they are fatal; some are so mild that you wouldn’t notice that a person was having one. The symptoms that result from a seizure are very individualized and depend on what is causing the seizure in the first place. You may wonder what these different types of seizures all have in common. The answer is that they are all characterized by out-of-control electrical activity in the brain. Wherever this activity is present it derails the normal functions of the brain. The other thing all seizures have in common is that you don’t want to have them. At best they disrupt your life and at worst they are very dangerous.

Wilder PenfieldFor most people with epilepsy, seizures can be controlled with medication, but for some, medication doesn’t do the trick. If this is the case and if the epilepsy is focal (meaning that it starts in one part of the brain), then there is another treatment option. If a small part of the brain is causing the problem, then the problem can be fixed by taking out that part of the brain. This may sound a little medieval, but the technique was actually developed in the 1930s by a neurosurgeon named Wilder Penfield and a neuroscientist named Herbert Jasper. The surgical procedure is called the Montreal Procedure because it was developed in Montreal, which shows that neuroscientists can be creative in some ways and less creative in others. The objective of the Montreal Procedure is to precisely determine which part of the brain is causing the epilepsy. This is important because the brain doesn’t have any spare parts. The popular myth that you only use 10% of your brain isn’t true. The whole brain is important, and you don’t want to lose any more of it than is absolutely necessary. You can figure out approximately where seizures are coming from by recording the electrical activity of the brain from the scalp (an electroencephalogram, or EEG). But, since this is brain surgery, its better to get closer than a ballpark estimate. So, in the Montreal Procedure the patient is kept awake while the surgeon opens up the skull. Once the brain is exposed the surgeon uses an electrical probe to stimulate different locations on the surface of the brain. It doesn’t hurt because the brain has no pain sensors. Now here’s the good part: when the surgeon stimulates the brain the patient can report the effect of the stimulation because he is awake. Depending on the location of the stimulation, the patient might make a movement or recall a memory or (you guessed it) feel something touching a part of his body. This allows the surgeon to identify the seizure focus, which is why you do it. There is a side bonus in that it also allowed Penfield to make functional maps of the cortex. This is how we know much of what we know about how the motor and sensory parts of the cortex are organized.

Because of Wilder Penfield’s epilepsy surgeries we know that stimulating a particular part of the brain (the primary sensory cortex) evokes the sensation of touch. Could we use this fact to restore the sense of touch after SCI? That is the question, but we need some more information before we can answer it. We need some more data, which means we need to electrically stimulate some people’s brains. Of course, opening up people’s skulls and electrically stimulating their brains for the sake of science is not an ethical thing to do. Fortunately, there is a group of people who are having their brains stimulated anyway, because surgery is still the best treatment option for many epilepsy patients. So, in order to make any headway on this issue we need a neural engineer who is also an epilepsy neurosurgeon. Jeff Ojemann is just such a person and he is tackling this very problem. In my next post I will tell you how.