Communicating Through a Vein in the Brain
Heart surgery used to be a Big Deal, and for good reason: accessing the heart required difficult and invasive surgery. Today, many heart operations can be done under nothing more than a local anaesthetic, during which a stent is inserted through an artery in the wrist or groin and guided into its final position close to the heart. This less invasive surgery has made the procedure safer, faster, and cheaper.
It appears that intracranial recording of brain activity is moving in the same direction. To measure high-quality neural signals, electrodes are typically inserted into the brain through small holes in the skull. A new technology, dubbed the “Stentrode”, offers a much less invasive way of implanting electrodes. Similar to a heart stent, the sensors are fed up into the brain through a large blood vessel — in this case, the jugular vein in the neck. This might just be the first intracranial brain-computer interface (BCI) designed for everyday home use, enabling paralysed patients to regain independence by controlling a laptop or mobile phone using their thoughts.
The communication problem
Diseases that cause motor impairments affect millions of people across the world. Whether they result from a sudden event like a car crash or a stroke or are the result of a more gradual degenerative disease like ALS, the impact on people’s daily lives is huge. People who recently led active and independent lives find themselves restricted, often requiring help with daily tasks like dressing, washing, and eating. But humans are social creatures, and one of the most challenging aspects of these conditions is their reduced ability to communicate with friends and family.
Patients with advanced ALS find it difficult to speak as they have limited control of their facial muscles. Yet many years before this, reduced dexterity in the hands and fingers may rob them of the ability to use a conventional computer or smartphone, leaving them dependent on caregivers for relatively banal — and yet socially hugely important — activities, such as sending a text message or buying items online. BCIs have the potential to radically improve independence for people in this situation, which can have a huge impact on personal quality of life.
Accuracy versus safety
The goal of any BCI is to transmit information directly between the brain and a computer. For example, a BCI user might want to communicate by typing out a text message or control the movement of a prosthetic limb. Whilst BCI applications are varied, in all cases the computer first needs to detect what’s going on inside the brain (sensing) and then interpret the user’s intention from this data (decoding).
To achieve high data quality from any sensor, it is important to take measurements from as close to the signal source as possible. We all know that our wi-fi signal is strongest closer to the router and that microphones work best when they are held close to the person speaking. In the same way, a BCI sensor is ideally located as close as possible to the source of the signals it is measuring — in this case, neurons inside the brain.
It follows that the most accurate neural recordings require the use of implantable sensors, such as the Utah array or the Neuralink device. However, these will always entail risks associated with invasive surgeries: a chance of infection, long recovery times, and the potential to rupture one of the many blood vessels covering the surface of the brain. The alternative is to use non-invasive sensors such as EEG or fNIRS, which are safer as they sit outside of the skull. Unfortunately, these are also much more limited in terms of data quality as neural signals are absorbed as they propagate through the cerebrospinal fluid, skull, and skin that protect the brain. Hence, there is a fundamental trade-off between accuracy and safety of the recording devices, and whilst advanced software approaches like AI can maximise the features that we can extract from the data, there is no way of recovering information that has simply been lost.
In practice, the choice of invasive versus non-invasive brain recording comes down to the intended use of the device. For consumer applications, cost and safety are critical, and the answer is almost always to use EEG sensors (as seen in Muse, Emotiv, and many other wellness or productivity-focussed consumer neurotech devices). But when the goal is to reliably control a sophisticated prosthesis, high-quality data becomes essential, and invasive sensors become a necessity.
Synchron’s approach
Research groups around the world are developing invasive BCI technologies for medical applications, from microelectrode arrays that control advanced prostheses, to ECoG implants for communications, to deep brain electrode stimulation to reduce Parkinson’s tremor. Advances over the last 10 years have been impressive, but all these devices share one limitation: the electrodes need to be placed in direct contact with the brain through a hole in the skull, meaning that invasive surgery is required.
This is where Synchron has taken a different approach. As Nick Opie, CTO of Synchron and co-inventor of the Stentrode technology, explains: “We saw an opportunity to develop a technology that could access the brain to extract high-quality information, while also ensuring that the surgical risks associated with performing a craniotomy were mitigated. Due to our technology being minimally invasive in nature, implantation of the Stentrode can be achieved in a day procedure, taking only a few hours.”
The Stentrode itself is a tiny mesh, or net, which is tightly folded inside a catheter — a small tube only 1 mm in diameter. The catheter is inserted into a vein in the neck, from where it travels up into the blood vessels that feed the brain. Here the mesh is deployed so that it sits around the edge of the blood vessel, with 16 tiny sensors monitoring the electrical signals of the motor cortex from inside the brain itself.
Synchron has recently published results from their first-in-human trials of the Stentrode, which focussed on improving communication for people with motor impairments. Two patients suffering from ALS, a form of progressive paralysis, had devices implanted and used the technology to regain some independence by directly controlling a computer without using their hands or their voice. In this first demonstration of the device, following an initial period of training, the Stentrode has been able to reliably distinguish between 3 states: “do nothing”, “click”, and “zoom”, with over 90% accuracy. The user makes these selections by imagining different physical actions such as clenching their fist or tapping their foot. Used together with an eye tracker, this combination allows the user to quickly make selections on a screen and intuitively interact with a conventional computer or mobile phone interface.
“This technology will enable people with any form of paralysis to use computers and assistive technology with their minds,” says Opie, “restoring their ability to communicate with their families and community; bank, shop, and use online resources; and enhance their quality of life through restoring mobility and independence.”
How does the Stentrode compare to other BCIs?
The Stentrode offers some significant advantages — but also some disadvantages — compared to other BCI technologies. It is somewhat of a halfway house: data quality is significantly higher than what can be achieved with non-invasive EEG, but still a lot lower than an electrode array implanted directly into the cortex. Installing the device still requires a surgical procedure, but the risks are reduced as this can be done via a blood vessel instead of through the skull.
What sets the Stentrode apart is that, unlike many of the more research-focused BCI systems, Synchron has designed their device to be fully implantable. The device is powered remotely, and data is transmitted wirelessly to a receiver outside of the body. This makes the system straightforward for a carer to set up, so that it can be used at home on a daily basis. However, this wireless method limits the maximum achievable data transfer rates, compared to high transfer rate systems like the BrainGate, making the Stentrode more appropriate for binary outputs like “click” and “zoom” as opposed to continuous actions such as controlling the precise movement of a prosthetic limb.
An additional inherent limitation to the technology is that the Stentrode sensors are located inside cerebral blood vessels, restricting the brain areas that it can record from to those adjacent to the larger arteries and veins.
Impact
It is no exaggeration to say that having the Stentrode fitted has proven life-changing for Synchron’s first study participant and his family. He is now able to browse the internet, control his own finances, and send text messages independently. He is even writing a book — “Technopathy for Beginners” — using this new technology. Implantation of his Stentrode has no doubt greatly enhanced his personal quality of life, and this effect ripples even further.
Opie describes how “surprisingly, my proudest moment was actually when his wife and carer informed us that she was now able to live her life independently, as she could now venture outside to the garden and go shopping knowing that he was able to contact her and communicate with her using text messaging applications that we set up and trained him to use. Knowing that our work impacted her as well as the recipient was very humbling.”
Synchron is working hard to get this technology into the hands — or rather, brains — of all those who can benefit from it. They have already begun human clinical trials and are planning to launch their first product within the next 5 years. But beyond this, Opie says, “Synchron is really just beginning. We have almost unlimited vascular access to different brain regions with the technology and we will begin developing technologies that can treat other neurological conditions that currently require invasive brain surgery for implantation of electrodes — including Parkinson’s disease, epilepsy, and depression.” It seems likely that minimally invasive devices like the Stentrode will transform the field of BCIs, much like keyhole surgery has transformed cardiology.
Written by Dr. Hannah Claridge, edited by Simon Geukes, Suhela Kapoor and Cris Micheli, with artwork by Firas Safieddine.
Dr. Hannah Claridge leads the Neurotechnology team at TTP, Europe’s largest independent deep tech consultancy. Hannah is a product development physicist and has developed clinical technologies for household names as well as contributing to the success of ambitious start-ups. With a Master’s degree in Physics and a Ph.D. in Clinical Neurosciences from the University of Oxford, she has a particular interest in biosensing and neuromodulation therapies.
Simon Guekes recently graduated from the University of Amsterdam, where he investigated brain-computer interfaces and the sensorimotor cortex.
Suhela Kapoor is a neurobiologist managing her digital health startup in India.
Cris Micheli, PhD is a senior software engineer, BCI / biosignals researcher and project manager with Cognixion.
Firas Safieddineis a Barcelona-based designer, architect, artist, researcher, and neurotech enthusiast.
