Researchers at the University of Melbourne and Royal Melbourne Hospital have reported successful results from a clinical trial of a brain-computer interface that enables communication for patients with severe paralysis. Four participants with motor neurone disease used the system to control a computer cursor and type messages through thought alone.

The system, called Stentrode, differs from most brain-computer interfaces by avoiding open brain surgery. Instead, surgeons insert the recording electrodes through blood vessels, threading them to the motor cortex. Once in position, the device records neural signals and transmits them wirelessly to external computers.

How It Works

The Stentrode device resembles a small stent, about the size of a paperclip. It contains 16 electrodes arranged around the circumference of the device. After being inserted through a blood vessel in the neck, it self-expands in a blood vessel on the brain’s surface, positioning the electrodes against the vessel wall where they can detect neural activity.

When a patient thinks about moving, neurons in the motor cortex fire in distinctive patterns. The Stentrode electrodes pick up those patterns and transmit the data wirelessly to a receiver worn on the body. Machine learning algorithms then interpret the neural signals and translate them into cursor movements or typed commands.

The trial participants underwent training to optimise their control. Initially, the algorithms had to learn each person’s unique neural patterns. Then patients practiced thinking specific movement intentions while watching the cursor respond. Over several weeks, most participants achieved typing speeds of 15-20 characters per minute.

That’s slower than typing on a keyboard, but it’s substantially faster than many assistive technologies available to people with severe paralysis. Eye-tracking systems, which many motor neurone disease patients currently use, typically achieve 5-10 characters per minute under ideal conditions.

The Clinical Trial

The trial enrolled four patients with advanced motor neurone disease who had lost the ability to communicate effectively through speech or typing. All had intact cognitive function but almost complete paralysis. For these patients, brain-computer interfaces offer one of the few remaining options for autonomous communication.

Safety was the trial’s primary endpoint. The concern with any neural implant is infection, immune responses, or damage to brain tissue. The endovascular approach used by Stentrode aims to reduce those risks by avoiding direct contact with brain tissue and requiring only a small incision for insertion.

All four participants successfully received the implant without serious complications. Three achieved reliable cursor control within two months. The fourth participant, who had more advanced disease progression, achieved limited control but continued to use the system for communication.

Participants reported substantial improvements in quality of life. Being able to compose messages independently, rather than relying on caregivers to interpret eye movements or blinks, restored a degree of autonomy and privacy. Several participants used the system to write emails, browse the web, and maintain social connections.

Comparison to Other Approaches

Several groups worldwide are developing brain-computer interfaces, using different technical approaches. Elon Musk’s Neuralink uses flexible electrode arrays implanted directly on the brain’s surface through skull openings. While potentially offering higher-quality neural recordings, that approach requires more invasive surgery.

The BrainGate consortium at Brown University has demonstrated remarkable cursor control and even robotic arm manipulation using arrays of micro-electrodes that penetrate into brain tissue. Those systems achieve faster typing speeds than Stentrode, but again require open brain surgery.

The Melbourne team argues that their endovascular approach offers the right balance for many patients. The surgery is relatively low-risk and can be performed under local anaesthetic. While signal quality doesn’t match penetrating electrode arrays, it’s sufficient for useful cursor control.

There’s also the question of longevity. Penetrating electrodes often lose signal quality over time as tissue reactions insulate them from neurons. The Stentrode approach may avoid some of those problems by not penetrating tissue. The trial participants maintain good signal quality more than a year after implantation, but longer-term data is still accumulating.

Commercialisation Path

Synchron, the company commercialising Stentrode technology, has expanded trials to the United States with FDA approval. The company aims to gather enough safety and efficacy data to seek full regulatory approval within a few years.

If approved, the system would initially target patients with motor neurone disease and other conditions causing complete paralysis. Later applications might include stroke patients, spinal cord injuries, and potentially other neurological conditions.

Pricing and reimbursement remain uncertain. Medical devices requiring surgery and ongoing technical support are expensive. Health systems will need convincing evidence that the quality-of-life improvements justify the costs. That calculation differs substantially between countries with different healthcare funding models.

The Australian trial participants have provided valuable feedback on practical aspects of living with the system. Battery life for the wireless transmitter, software reliability, and how the system integrates into daily routines all matter for real-world usability. Those insights are informing ongoing device development.

Broader Context

Brain-computer interfaces have been a research topic for decades, but only recently have they transitioned from laboratory demonstrations to clinical applications. Improved understanding of neural coding, better signal processing algorithms, and miniaturised wireless electronics have all contributed.

Australian researchers have been notably active in this field. The Bionic Vision Australia consortium developed retinal implants for vision restoration. The University of Melbourne has groups working on cochlear implant improvements. There’s substantial expertise in implantable medical devices and neural engineering.

That expertise builds on Australia’s history in biomedical device development, particularly cochlear implants commercialised by Cochlear Ltd. The skills and supply chains developed for those devices translate reasonably well to other neural interfaces.

For patients with paralysis, any technology that restores communication or control represents profound improvement. The Stentrode trial results suggest that safe, practical brain-computer interfaces are moving from research concepts to clinical reality. The path to widespread availability remains long, but the direction is clear.