Adrian Watkins

It was 2012, and research into brain-machine interfaces was stuck. The technology, which links brain waves to devices such as computers or robotic arms, promised to open the world back up for paralyzed people—but it was held up by a fundamental issue. The surgery to implant a sensor involved drilling into people’s skulls.

Several companies have sprung up around this technology, including BrainGate and Elon Musk’s recent entry, Neuralink. But the high bar around approaches that require open-brain surgery means the journey to commercialization is long and tough.

“That’s part of the reason we got to human trials first,” said Thomas Oxley, M.D., Ph.D., CEO of Synchron, a company that puts its sensor into the brain through blood vessels using a catheter. Oxley pitched the idea to the Defense Advanced Research Projects Agency, and the company, then called SmartStent, was born.

The company has implanted its Stentrode device into the first patient of a clinical trial in Synchro’s native Melbourne, Australia. The study will assess the technology’s ability to restore communication in people with severe paralysis caused by various conditions, including stroke, spinal cord injury, muscular dystrophy or amyotrophic lateral sclerosis.

The Stentrode, along with a transmitter implanted in the chest and a set of “cross-platform” apps, form a brain-machine interface that captures brain signals and translates them into actions.

Like its name suggests, the Stentrode looks like a stent—the thin devices used for decades to treat cardiovascular diseases—but with sensors on it. It is delivered through the jugular vein up into the brain and is deposited in the motor cortex. Although this part of the brain is often called the “command center” that controls the body’s movement, Synchron thinks of it more as a “volition center.”

“Not movement, but the control of movement. We want to take the thoughts or the free will of a person who wants to attempt to move something and turn it into a digital output,” Oxley said.

During the implantation procedure, the physician runs a wire down from the brain under the skin to the chest. Captured brain signals travel down this lead to a second implant, called BrainPort, which transmits the data wirelessly to a smart device that processes it and interacts with other devices running Synchron’s software.

All of this takes place through a 2-mm incision in the neck, Oxley said.

Synchron thinks about overcoming paralysis in three ways: “failure of talking, failure of your hands and failure of your legs,” Oxley said. “We want to test the apps we are building in a trial to demonstrate that patients can control an app that overcomes paralysis, starting with speech.”

Most people think of speech as talking out loud, with Stephen Hawking’s talking machine coming to mind. In contrast, Synchron is focusing on the digital word, in Oxley’s words: texting.

“We are essentially building a UX that enables people to deliver text messages completely hands-free—that’s what we consider digital speech. We are generating key presses that a patient learns to control through our UX that we’ve recorded through the motor cortex,” he said. The study will look at how fast and how accurately patients write text messages using the interface.

Synchron’s system uses electrocorticography, or ECoG, to measure electrical activity in the brain. It then uses machine-learning algorithms to break the brain signal down into different frequencies and connect particular thoughts with spikes in different frequencies.

“The algorithm knows when a user is attempting to do something and watches for the spikes in frequency changes and map that to an output. The patient learns to control the process,” Oxley said. “We’re taking a part of the brain that previously controlled muscle and, now the patient is paralyzed, we are repurposing that part of the brain to control a digital output.”

Synchron is in talks with the FDA over its regulatory strategy and plans to use data gleaned from this first study to design a pivotal study for U.S. approval.

Full article here: