Neuralink is building a fully integrated brain machine interface (BMI) system. Sometimes you'll see this called a brain computer interface (BCI). Either way, BMIs are technologies that enable a computer or other digital device to communicate directly with the brain. For example, through information readout from the brain, a person with paralysis can control a computer mouse or keyboard. Or, information can be written back into the brain, for example to restore the sense of touch. Our goal is to build a system with at least two orders of magnitude more communication channels (electrodes) than current clinically-approved devices. This system needs to be safe, it must have fully wireless communication through the skin, and it has to be ready for patients to take home and use on their own. Our device, called the Link, will be able to record from 1024 electrodes and is designed to meet these criteria.

Neuralink’s technology builds on decades of BMI research in academic labs, including several ongoing studies with human participants. The BMI systems used in these studies have no more than a few hundred electrodes, with connectors that pass through the skin. Also, their use requires laboratory equipment and personnel. Our challenge is to scale up the number of electrodes while also building a safe and effective clinical system that users can take home and operate by themselves. Recent engineering advances in the field and new technologies developed at Neuralink are paving the way for progress on each of the key technical hurdles.

ELECTRODES

In order to optimize the compatibility of our threads with the surrounding tissue, they should be on the same size scale as neighboring neurons and as flexible as possible. Therefore, we microfabricate the threads out of thin film metals and polymers. But the threads also have to resist corrosion from fluid in the tissue, and the electrodes must have sufficient surface area to allow stimulation. To meet these criteria, we’ve developed new microfabrication processes and made advances in materials science. These include the integration of corrosion-resistant adhesion layers to the threads and rough electrode materials that increase their effective surface area without increasing their size.

CHIPS

Our Link needs to convert the small electrical signals recorded by each electrode into real-time neural information. Since the neural signals in the brain are small (microvolts), Link must have high-performance signal amplifiers and digitizers. Also, as the number of electrodes increases, these raw digital signals become too much information to upload with low power devices. So scaling our devices requires on-chip, real-time identification and characterization of neural spikes. Our custom chips on the Link meet these goals, while radically reducing per-channel chip size and power consumption over current technology.

HERMETIC PACKAGING

The Link needs to be protected from the fluid and salts that bathe surrounding tissue. Making a water-proof enclosure can be hard, but it’s very hard when that enclosure must be constructed from biocompatible materials, replace the skull structurally, and allow over a 1,000 electrical channels to pass through it. To meet this challenge, we are developing innovative techniques to build and seal each major component of the package. For example, by replacing the connection of multiple components with a process that builds them as a single component, we can decrease device size and eliminate a potential failure point.

NEUROSURGERY

Our threads are too fine to be manipulated by hand and too flexible to go into the brain on their own (imagine trying to sew a button with thread but no needle). Yet we need to safely insert them with precision and efficiency. Our solution is based on a new kind of surgical robot, whose initial prototype was developed at the University of California. We are innovating on robot design, imaging systems, and software, to build a robot that can precisely and efficiently insert many threads through a single 8 mm skull opening while avoiding blood vessels on the surface of the brain.

NEURAL DECODING

Neural spikes contain a lot of information, but that information has to be decoded in order to use it for controlling a computer. Academic labs have designed computer algorithms controlling a virtual computer mouse from the activity of hundreds of neurons. Our devices will be able to connect to over an order of magnitude more neurons. We want to use the additional information for more precise and naturalistic control and to include additional virtual devices such as a keyboard and game controller. To accomplish this, we are building on recent advances in statistics and algorithm design. One challenge is to design adaptive algorithms that maintain reliable and robust performance while continuing to improve over time, including the addition of new capabilities. Ultimately, we want these algorithms to run in real time on our low-power devices.

There are currently only a few approved devices for recording or stimulating from the human brain, including devices for deep brain stimulation (DBS), which can treat neurological disorders such as Parkinson’s Disease, and devices for the detection and disruption of seizures. These are designed to modulate the activity of whole brain areas, not to transfer information to and from the brain. Therefore, they only have a small number of electrodes (less than 10) and are much larger than our threads. For example, DBS leads have only 4-8 electrodes and are about 800 times larger.

There are other devices being used in clinical trials for BMI movement control or sensory restoration. However, none of these devices have more than a few hundred electrodes, and they are all either placed on the surface of the brain or in fixed arrays of single rigid electrodes. The Link is being designed with an order of magnitude more electrodes and with flexible threads that are individually placed to avoid blood vessels and to best cover the brain region of interest.

We are also designing the Link to provide unprecedented scale, with over 1024 channels of information from the brain. The Link will also perform real-time spike detection on every channel, and these data will all be sent wirelessly.