Neuralink PRIME Study

PRIME is Neuralink’s first-in-human study of the N1 implant and the R1 robot. The letters, as the company expands them, stand for Precise Robotically Implanted Brain-Computer Interface. The devices are investigational. Neuralink’s own posts say it does not guarantee a benefit from taking part.

What the registry says

NCT06429735 lists Neuralink Corp as sponsor. The official title is an early feasibility study of a robotically implanted brain-computer interface for control of external devices. The record describes a skull-mounted wireless rechargeable implant whose threads are inserted by the R1 robot, in participants with tetraparesis or tetraplegia. When this page was checked, the record showed a January 9, 2024 start, recruiting status, estimated enrollment of 15, and a primary completion of June 2026. Registry fields move. Quote the live record, not this sentence, if the numbers matter.

A company brochure linked from Neuralink describes a primary phase of about 18 months and a long-term follow-up of about five years. That brochure is a recruitment document, not an outcome paper.

What the company has reported

The September 19, 2023 note says an institutional review board and the first hospital had cleared recruitment, under an FDA investigational device exemption the company dates to May 2023. People with quadriplegia from cervical spinal-cord injury or ALS were the group named.

The April 12, 2024 note says the first implant was placed in January 2024 at Barrow Neurological Institute in Phoenix, that neural signals were detected soon after, and that the participant used the system for applications including online chess. The company names the surgical target as the precentral gyrus, chosen with functional MRI. It describes the operation as a neurosurgeon opening the site, the robot inserting threads, and the implant body sitting in the opening under the scalp.

The May 8, 2024 note names that participant, Noland Arbaugh, and is the place to read the limits:

• Use is company-reported: cursor control, games, a laptop while lying down, a week that included 69 hours of use.
• Performance is company-reported on Neuralink’s grid task: 4.6 bits per second on the first research session, later 8.0.
• Threads retracted in the weeks after surgery. Effective electrodes dropped, then software changes brought the score back. That is a company account of a hardware problem, not a claim that the issue cannot recur.

The August 21, 2024 note names a second participant, Alex, also implanted at Barrow and discharged the next day. The company says he was moving a cursor in less than five minutes and, on day two, used CAD software. It says the mitigations for retraction were reducing brain motion during surgery and closing the gap between the implant and the brain, and that no retraction had been seen in his case at the time of writing.

Related studies, not the same study

CONVOY, NCT06710626, is a separate early feasibility record for people already in PRIME and already implanted, invited to try controlling assistive devices including a robotic arm. Estimated enrollment on the record is 3. It is not evidence that robotic-arm control already works in daily life.

GB-PRIME, NCT07127172, is the U.K. feasibility record, with sites at Newcastle upon Tyne Hospitals and University College London Hospitals. Estimated enrollment on the record is 7.

What this page will not say

Playing a game on a company blog is not a peer-reviewed efficacy result. Breakthrough Device Designation for Blindsight, announced September 17, 2024, is a different device and is not permission to market an implant. “Restoring autonomy” is the company’s aim. The study’s job, as registered, is to look at initial safety and whether the device functions.

Sources

• Recruitment announcement, September 19, 2023
• Progress update, April 12, 2024
• User experience, May 8, 2024
• Second participant, August 21, 2024
• PRIME study brochure (PDF)
• NCT06429735
• NCT06710626
• NCT07127172

Related

• Neuralink
• Companies
• Receipts
• Risk tolerance