Neuralink gets the headlines for brain implants. The first brain-computer interface approved for ordinary commercial sale is not Neuralink's, is not American, and — the part most coverage underplayed — does not go into brain tissue at all.
China's National Medical Products Administration granted market approval to NEO, developed by Shanghai-based Neuracle Medical Technology with researchers at Tsinghua University, while Neuralink remains in human trials.
Who it is for, and what it does
NEO is approved for patients aged 18 to 60 with quadriplegia caused by cervical spinal cord injury. A coin-sized implant with eight electrodes collects and decodes brain signals in real time, and the patient uses them to drive a pneumatic glove that grips and releases.
The goal is restoring a specific lost function — picking up a cup, holding a spoon — not a general-purpose interface. That modesty is the design, not a limitation of it.
The difference from Neuralink, stated precisely
The distinction is where the electrodes sit. Neuralink threads electrodes into the cortex, the brain tissue itself, which yields the richest signal and the highest surgical risk. NEO's electrodes rest on the dura mater — the tough protective membrane around the brain — and never enter the tissue beneath. Signals are read through the membrane and sent to a computer that translates them into commands.
The implantation takes roughly 90 minutes. It is widely assumed that this lower surgical risk is what persuaded regulators to approve NEO for general sale rather than restrict it to trials.
The engineering trade-off is permanent and worth stating plainly. Deeper electrodes read more and damage more. NEO's bet is that for the specific job of restoring hand movement, a lower-resolution signal from outside the brain is good enough. Neuralink's bet is the opposite: that richer signals unlock capabilities — cursor control, speech — that a surface device never will. Both can be right for different patients. For now, only one is on sale.
The detail that will decide whether it reaches anyone
China has added NEO to its national health insurance system. For patients this is arguably more consequential than the approval.
An approved implant that is not reimbursed is a device for the wealthy and for clinical trials. Spinal cord injury disproportionately affects working-age people who lose their income at the moment of injury, and the population that needs an assistive implant is substantially the population least able to buy one. A regulator can make a device legal; only a payer makes it available.
That is the part health systems elsewhere will be watching, and it is a reasonable question to ask of every neurotech approval that follows: approved, yes — but reimbursed by whom?
What the evidence can and cannot show yet
NEO has been through clinical trials on several dozen patients. Trials of that size can demonstrate that a device works and is safe enough to approve. They cannot show how it performs across thousands of patients over years — electrode drift, signal degradation, infection rates, how long the implant keeps working. The commercial rollout is where that gets answered, which is the ordinary and slightly uncomfortable position of every newly approved implant.
Sources: China's National Medical Products Administration; Neuracle Medical Technology; Tsinghua University; Macau Business; Galaxus; and Prothom Alo's original report.
