What Will It Take to Design a Brain Implant That Lasts a Lifetime?

When Ian Burkhart was just 19 years old, he lost the ability to move below his elbows after a diving accident. Three years later, he came across a clinical trial that would begin in 2013 and could potentially “reanimate” him. This medical miracle, he was told, would come at a cost. Researchers at the Ohio State University would implant a small brain-computer interface in Burkhart’s head. He said the conversation with his family and friends wasn’t easy. “It was like, ‘You just had a spinal cord injury three years ago, and you lost mostly everything, and you still have your brain. Why do you want to risk damaging your brain now?’”

Burkhart took that risk. The implant was a success, and for several years, he lived a curious double life. At home, he was still paralyzed. But in the lab, after switches flicked and sensors roused to life, the electrode array mounted on the surface of his brain ran signals buzzing with intention to a pedestal unit screwed into his skull, and then to an external computer. This relayed the impulses to a sleeve-like stimulator on his arm, which delivered the electricity that could no longer pass directly through Burkhart’s spinal cord to his muscles. In the lab, he could pour a glass of water. He could swipe a credit card. He could play Guitar Hero.

Burkhart became paralyzed for the second time in the summer of 2021, when he asked the trial team to remove the device. The trial, originally slated to last roughly 18 months, had been extended repeatedly. But the academic institutions that had installed the device were running low on funds, said Burkhart. Additionally, the open wound in his head that had remained from his operation developed a nagging infection. “I'd much rather have this device be explanted as part of a scheduled operation versus [when] infection is raging through my system,” he explained.

Burkhart’s dice throw on a chance at restored function is one every brain-computer interface pioneer has taken since neurologist Philip Kennedy implanted the first in-human brain-computer interface in 1998. To date, fewer than 100 people have received a brain-computer interface implant.

Making these devices widely available to the public will require commercial brain-computer interface companies to offer a much surer bet than the one Burkhart accepted. By exploring novel materials and engineering new configurations, academic and industry scientists are developing new brain-computer interface technology that they hope will address this challenge. They aim to enable future devices to last well over a decade without requiring repeated invasive brain surgery and that these new implants will weather financial and technological change. Producing a device that’s up to the task will require mastery of the brain and of bionics.

Designing a Brain-Proof Electrode for a Long-Lasting Brain-Computer Interface

Burkhart’s brain-computer interface was a Utah array—a device as close as the nascent field of brain implants has to a reliable workhorse.¹ The array, a 4.2mm square grid studded with 100 silicon microelectrodes, has been implanted in dozens of patients over decades of testing. The Utah array’s electrode tips burrow into the cortex to pick up electrical signals. But the brain is a hostile environment that can wear down electrodes, said Kurt Haggstrom, the chief commercial officer of the implant company Synchron. Developing a brain implant that can endure and perform in these conditions is far from straightforward.

One challenge is electrode retraction, in which shifting brain tissue can break free of the spines on the electrodes, reducing connection strength. Sometimes, the signal from an electrode array isn’t usable by the BCI, or it can drop off without explanation. “That's why it's important to have as many electrodes as you can,” said Matt Willsey, a neurosurgeon at the University of Michigan. But the most significant barrier to long-term brain-computer interface stability is the brain’s in-house police force, immune cells called microglia.

“Your body is made to not like invaders,” said Michelle Patrick-Krueger, a neuroscientist at the University of Texas who published a review of brain-computer interface trials earlier this year.² Brain-computer interface implantation awakens nearby microglia, which, in concert with other non-neuronal brain cells such as astrocytes, sheath the implant, reducing its connection to nearby neurons.³ Microglia release additional pro-inflammatory molecules and reactive oxygen species, which oxidize the electrode’s surface and ultimately corrode the implant. This corrosion can reduce or impede the transmission of electrical signals between neurons and electrodes.⁴ Analysis of explanted older brain-computer interfaces consistently shows a buildup of physical damage and a steady reduction in signal quality over time.⁵ Nevertheless, the goal for brain-computer interface designers is to minimize or avoid this immune response.

Small, Soft, or Stiff? Choosing Brain-Computer Interface Materials

Modern electrode arrays are made from strong, inert metals such as iridium to prevent electrode degradation and maximize signal strength. But these implant materials, which also include silicon and glass, are thousands of times stiffer than the brain’s surface. Much research has gone into developing conducting materials that more closely resemble the soft brain. The leading paradigm, said Niclas Roxhed, who studies biomedical microsystems at the KTH Royal Institute of Technology, is, “If you want to put something in a soft brain tissue, your implant also needs to be soft.”

Still, Roxhed recently published a paper suggesting otherwise.⁶ His team developed a silicon neural implant for drug delivery, roughly the width of a single hair. By going small, rather than soft, Roxhed’s device avoided triggering extensive immune responses in preclinical trials in mice.

Commercial brain-computer interface companies are following this route as well. Former Stanford University researcher Matt Angle is the chief executive officer of the Austin, Texas-based brain-computer interface company Paradromics. He agreed with Roxhed’s assessment that size matters in brain-computer interface design. “It's one of these brain myths that electrodes scar because they're too stiff,” said Angle. He maintains that miniaturization is the key to bypassing overactive immune responses.

Paradromics’ Connexus brain-computer interface was inspired by the Utah array but is far smaller; the electrode’s spines are 40 microns in diameter, less than half the size of the Utah array's. The result, said Angle, is a reduced inflammatory response. In unpublished internal company data, Paradromics said the Connexus has demonstrated stable signal performance over 12 months of implantation in large-animal models.

The Connexus's first long-term clinical implantations will occur as part of Paradromics' Connect-One study. The brain-computer interface aims to restore speech in two patients with loss of voluntary motor control. Connexus’s key innovation is that it has a far higher information transfer rate than that of other commercial devices. Ensuring that brain immune responses don’t disrupt this signal will make or break the device’s speech translation ability, so the trial will be the real test of the Connexus's stiffer, smaller approach.

Synchron’s approach, in contrast, bypasses the brain almost entirely. Their Stentrode device, which combines a wire-mesh tube made from a nickel alloy with mounted platinum-iridium electrodes, is delivered into the body via the jugular vein and sits next to the motor cortex within a blood vessel. From this position, it can record nerve impulses and send motor intentions to a computer via a relay implanted in the chest. Synchron’s device has achieved impressive in-human results, helping paralyzed patients control a touchscreen cursor.⁷ Haggstrom said that his company’s approach will avoid the brain’s challenging immune environment. The chest-mounted section of the implant can be easily replaced. Nevertheless, Willsey said that for certain brain-computer interface goals, such as complex fine-motor control, direct access to the cortex is still required.

Both Angle and Haggstrom said that they aim for their technology to last for decades within their patients. The Stentrode cannot be removed once it has settled in its target vessel. Even if these devices overcome the biological and physical barriers to long-term implantation, they will also face economic challenges that risk shortening their lifespan.

How Will Patients Pay for a Brain-Computer Interface?

Brain-computer interfaces are entering a “translational” era, where technologies that excel and excite in the lab reckon with commercial reality, said Patrick-Krueger. “This is going to show you what's going to work…what’s going to become viable.”

As more companies enter the field, the peril of some device manufacturers going under, leaving patients with functionless hunks of metal in their heads, increases. The pace of change in the field also risks leaving early adopters using obsolete technology for decades to come. The issue, said Angle, is that the financial model most appropriate for BCIs doesn’t exist yet. “[Brain-computer interface] would benefit from having both those obvious one-time device payments where the company can make a good profit margin on the original device, but also a recurring revenue stream,” he said.

Ongoing payment, Angle was quick to add, wouldn’t come from the patient—the Black Mirror-lite suggestion of a brain-computer interface subscription that can turn off a person’s hippocampus after a missed payment should remain a dystopian fantasy. A financial solution will be needed as the device must remain functional for a long time. The brain-computer interface devices will also require regular software updates to make sure users aren’t relying on outdated or incompatible technology. This digital element makes the task of maintaining a brain-computer interface more complicated than other implantable devices, for example, a pacemaker.

These essential questions, said Burkhart, are being hashed out by bodies like the iBCI Collaborative Community, a group of academics, brain-computer interface advocates, and company representatives convened by the Mass General Brigham health system, who are mapping out what the future of fair brain-computer interface use looks like.

Somewhere in that future, Burkhart said he’d like to once again benefit from a brain-computer interface .

But not yet.

“I want to wait for the next generation of devices,” he said. “I’ve already been through the big part of sacrificing my own body for science, and I am ready to benefit from the tech while pushing the boundaries of the next gen.”