The Future of Neuromodulation
University of Otago, New Zealand
Professor Dirk De Ridder’s main research topic is the understanding and treatment of phantom perceptions such as pain and tinnitus, as well as addiction, using non-invasive neuromodulation and especially invasive neuromodulation techniques such as brain implants.
The focus of his research is to understand the common mechanisms of different diseases such as pain, tinnitus, Parkinson’s Disease, depression, slow-wave epilepsy and a group of diseases known as ‘thalamocortical dysrhythmias’. His research also focuses on addiction, obsessive-compulsive disorder, impulsive and personality disorders, an entity called ‘reward deficiency syndromes’.
What is Non-Invasive Neuromodulation?
Non-invasive has been developed a long time ago because the Egyptians, the Greeks and the Romans used electrical fish, which they applied locally to the head, feet or back to treat pain or gout or some inflammatory problems. This fish was later replaced by batteries, once they developed in the 17th-18th century, that generate the electricity to drive either a magnetic stimulator, an electrical stimulator, or ultrasound stimulators and optogenetics (although these are currently only invasive). Non-invasive means that the electrical current or the magnetic pulses go through the skin and the skull into the brain. Still, you don’t need to open the head surgically in order to modify brain function and activity.
What is Invasive Neuromodulation?
Invasive neuromodulation is the surgical modulation where you insert electrodes in the brain, either on a peripheral nerve or in a peripheral part of the nervous system. The most well-known is Deep Brain Stimulation for movement disorders or Parkinson’s Disease (PD) where you can improve the movement problems that people with PD have predominantly by blocking a small nucleus in the brain. Still, it’s not that only the small nucleus is responsible; it’s a whole network attached to that nucleus that changes. So that is the invasive neuromodulation, but there are other devices as well. For example, there are electrodes that can be put against or around the vagal nerve, the sacral nerves, there are spinal cord implants as well. So basically, invasive neuromodulation can be done at any part of our entire nervous system.
How Can Neuromodulation Be Used to
Construct Healthy Patterns of Neuronal Behaviour?
What we really need is to not destructively interfere with ongoing activity in the brain as we do right now. We have to teach the brain how to learn something and that is what we don’t do with the current neuromodulation. The brain doesn’t learn anything with the neural pace and with adaptive stimulation that exists in some of the neuromodulation devices. These are all just like the surface system in a car, it’s responsive but it does not learn anything.
If we want to really make a difference, where we do not only destroy abnormal activity but that we want to give an input of new and good activity, then we have to be able to get the brain to learn something. One way of doing that, of course, is by using artificial intelligence. Another way, in its simple form, is by reconditioning stimulation where you implant an electrode in the reward system in the brain and then you present something that you want the brain to learn.
Neuromodulation and IoT (Internet of Things)
How can neuromodulation and the IoT be used to treat the obesity pandemic?
People with obesity want their brain only to be interested in low-calorie food. Then whenever they see low-calorie food, they will get a reward in the brain and thus, the brain will learn and will start liking low-calorie food progressively by giving a reward whenever the person sees low-calorie food.
If we go one step further, for example, we could depend on our Google maps or another map program that will tell you: “Well, behind the corner there is a food shop”, so you already start stimulating in your brain to suppress that craving for food before you get there. So you get a more integrated stimulator in the environment, which means that you can already contextually go and adapt, but also learn by being exposed to the stimulus that makes you pathological. Crucially this will require that we integrate whatever we already have, but another important step is to decode what our brain picks up from the environment.
How Can AI & Neuromodulation Be Used to Treat Pain?
If we can pick up activity in the brain that is a signature for pain and the different components of pain, so both for the painfulness and the suffering, you can automatically adjust the stimulation to abnormal signatures of pain in your brain so that the neuromodulation becomes a closed-loop. Initially just a simple closed-loop, but ultimately in a later stage, when the brain can pick up the pattern that predicts pain will ensue, it can start preventing the pain from arising.
Therefore, we need artificial intelligence because we cannot pick up these complex patterns and brain activity, and to do so, we need help from artificial intelligence that will help us decode those signatures. Pain is one example, but of course a thought, a feeling, or any other aspect of human life should be decodable. This is something that we cannot do ourselves. We are smart, but clearly not smart enough to decode 86 billion neurons and the same amount of client cells, which creates so many possibilities that we as humans, even with good models, cannot extract the correct patterns from such a complex system.