How Gertrude the pig could change the world and what it really means for us
Writer: Tara Spasojevic
Editor: Ebani Dhawan
Artist: Sam Vladimirsky
On 28th August, Elon Musk and his company Neuralink live streamed an update to announce their progress on the ‘Link’, a brain-computer interface (BCI). A BCI is a device that converts brain signals into an output, which, in this case, will be expressed on a phone app. In the update, Musk suggested that the Link has applications in memory loss, blindness, paralysis, depression, anxiety and addiction. Initial clinical trials are targeting individuals with spinal cord injury such as tetraplegia. Consider controlling your phone without even lifting a finger, simply by thinking of what you want to do; it’s probably as close to Derren Brown as you’re going to get. The Link is better than magic. It’s science.
This is by no means a novel premise; experimentation with BCIs has been around since the 1970s. Non-invasive techniques such as electroencephalograms (EEGs) read generalised brain activity, and advancements like the NeuroGrid promise more accurate tracking of neural spikes. Deep brain stimulation (DBS) is an effective treatment in some neurological disorders, but it can’t read or write high bandwidth information, nor can it be used in normal daily life. BCIs have been used for patients with severe motor disorders to control wheelchairs or prosthetic limbs, for example neuromotor prostheses using BrainGate technology. However, these are limited by a small number of commands ‒ Neuralink promises much more.
How does it work?
Information transfer in the brain happens by way of electrical impulses that travel from the soma (neuron cell body) down the axon to the dendrites, where the signal is propagated across the synapse in the form of chemical neurotransmitters. An electrode inserted in the extracellular space in the proximity of the neuron can detect the membrane potential created by electrical impulses or spikes ‒ this is referred to as spike detection. There are about 86 billion neurons in the brain, which makes individual spikes hard to detect and pinpoint. The Link, however, does this at a single-spike resolution.
The tentatively named Link version 0.9 is 23 x 8 mm in size and has 1024 channels distributed over 96 electrode threads capable of high-fidelity neuronal activity recording which is analysed in real time. Data is transferred wirelessly via Bluetooth. It has all-day battery life that can be charged via an inductive charger that you put on your head as you sleep. Recharging your batteries overnight has never been more literal. In order to implant the device, a coin-shaped section of the skull is removed and replaced by the Link.
The shift from experimental to clinical application is one of Neuralink’s foremost achievements. Much of this is enabled by the specialised, high-precision surgical robot that is able to perform full installation autonomously in under an hour (general anaesthesia is not required). Electrodes are inserted 3-4 mm into the cerebral cortical surface at a rate of 6 threads per minute; penetration into deep brain processes is still not feasible. The surgical robot screens the brain to map out the best way to insert the electrodes in designated brain structures and avoid blood vessels. This is vital as it prevents bleeding and reduces the risk of an inflammatory reaction to a foreign object in the brain.
Materials science has played a key role in making Neuralink’s ambitions a reality. The 96 threads are made of an ultra-thin biocompatible polymer designed to mimic a neuron axon as closely as possible. At less than 5 microns in width, they are 20 times thinner than a strand of hair. These small dimensions would normally mean incredibly high resistance at the brain/ electrode interface, yet material engineering has allowed development of extra thin flexible wires that move with the brain, minimising damage.
Other BCIs require in-depth analysis to untangle the messy feedback of neuronal spiking and understand what information is actually being sent. Instead, the Link analyses and writes the data in real time. Spike detection is done in less than 900 nanoseconds, facilitated by the custom-designed integrated circuit.
On the whole, the Link needs to be a biocompatible, hermetically sealed package that can last in the inhospitable environment of the human brain for decades and without harm to the user. Within a year it has already been dramatically simplified; according to Musk, it is essentially “like a FitBit in the skull with tiny wires”.
Three little pigs
Current trials with live pigs are being done with apparent success. During the presentation, Musk stressed the measures being taken to ensure the pigs’ proper care and wellbeing (it is quite easy to keep pigs happy, apparently). No animals in the trial displayed observable behaviour changes. Ironically, two-month implant carrier Gertrude was rather disinterested in posing as Musk’s poster pig, but she came around eventually. Meanwhile, her better behaved counterpart Dorothy was more cooperative. Dorothy had carried the implant for two months, which was then removed, demonstrating the reversibility of the device. The similarity between pig and human skulls, and the ability to train pigs to do simple tasks, makes them a useful model. The positive outcome indicates that the Link will be robust for humans. In July, the team received Breakthrough Device Designation from the US Food and Drug Administration (FDA) and are keen to go forward with human implantation soon.
Writing a letter to your brain
The cortex is the outer layer of the brain, the ‘white matter’, and is involved in low-level processing of motor and sensory information. It is clear that like its predecessors, the Link can read neural spike activity, but the 1024 electrodes are all able to read and write. Any combination of electrodes can stimulate neurons to form arbitrary waveforms and in essence send information to the brain. If controlled properly, a single electrode has the capability to influence 1,000 to 10,000 neurons.
Where a tetraplegic would have once received information from their limbs, the same information can be transmitted by electrode stimulation. The information is the same, only the source is different. What’s more, Musk postulates that in the long term, full body motion could be restored by implanting a Link in the brain and another at the site of spinal cord injury, hence bypassing the missing connection. Even someone with a severed spine would be able to walk again.
Is Thinkpol the future?
Neuralink’s dreams easily draw comparisons to a slew of sci-fi concepts. A brain implant eerily brings about unbidden images of spy comedy Kingsman where a world cleanse experiment by a lisping Samuel L Jackson goes wrong. Only the rich could afford the life-saving implant, but his plans are foiled resulting in their heads popping off in a spectacular firework display of gory confetti. Of course, Musk’s Neuralink is not fitted with a self-destruct function, but it does beg the question of safety and security. The team assures that interactions with brain data will be fully encrypted and authenticated before they can be accessed and that security considerations are being taken in every step of development. If the wireless transmission could be compromised, then Orwellian dystopia would be too close for comfort.
If executed as projected, the Link device holds vast opportunity to enable fantastical feats. The Neuralink team shared their dreams which include (consensual) telepathy, Terminator-like super vision, pain control and the ability to download your memories to replay at a later date. Society probably isn’t ready to live in a Black Mirror reality and is unlikely to happen any time soon ‒ there is no release date yet in sight. While Musk dreams of an ‘AI symbiosis’, the reality is that the Link could be completely life-altering to so many with debilitating disorders or injuries with an estimated price point of a few thousand for the whole procedure. Yet the potential to misuse such a powerful tool is very real. How Neuralink will progress is still uncertain, even with a skilled team behind it. The brain is truly complex and much is unknown, but this appears to be the future of science whether we like it or not.