braingate 1000 days later close up
Braingate's 100 electrode array is tiny, but it opens up the entire world to paralyzed patients who can use it to communicate with computers.

The latest report from academia: when you hack someone’s brain to turn them into a cyborg, count on it lasting for a few years. Scientists developing Braingate, the revolutionary device that translates brain impulses into commands for computer cursors and robotic wheelchairs, recently announced that the cranial implant was still working in a patient 1000 days after it was installed. Paralyzed patients fitted with Braingate have electrodes hardwired into their neurons which allow them to connect to computers directly. They think about moving their body and a computer interprets that signal like the movements of a mouse or touchpad. A female patient known as S3 underwent five days of rigorous testing to see how well the brain-computer interface hardwired in her skull was functioning. While many of the electrodes had failed, the patient’s overall success with the device was still high when asked to move a cursor on a screen (91.3%). Watch examples of the tasks she performed in the video below. S3’s success means that patients fitted with brain implants like these may be able to go years between upgrades to their hardware, making the process that much more appealing. No doubt about it, Braingate is still in its early stages, but this technology is nothing short of extraordinary.

Brain-computer interfaces are fairly awe-inspiring on their own, but the longevity of Braingate is truly noteworthy. As mentioned in their report in the Journal of Neural Engineering, the scientists behind Braingate first implanted S3 with her device back in 2005. The ‘1000 days later’ study was conducted in 2008. According to Brown University (where most of the Braingate team is associated) S3 continues to participate in clinical trials, suggesting her implant is still functioning at some level now almost another three years later! Braingate team members won’t comment on the most current data on S3 (as this latest work hasn’t been published yet) but I highly suspect I’ll eventually be writing another article about this device with ‘2000 Days’ in the title.

Of course, even if S3’s brain implant has been functioning for six years it probably hasn’t been doing so at full capacity. During the 2008 study, scientists found that only about 41 of the electrodes in S3’s brain were providing meaningful signals. Originally she was implanted with a 4mm x 4mm array of 100 electrodes (96 of which connected successfully to some degree). That means that less than half of the array was still playing a meaningful role in the 1000 days trial. Will S3 continue to lose electrode connections as the implant ages? Perhaps. Yet her performance was still very high (91.3% success) after her signals were re-calibrated for the five day study. S3 was able to center a cursor inside shapes, and move it to random targets fairly well (both tasks you would want if you needed to replace a mouse with a brain-computer interface). Even with less electrodes functioning, S3 was doing well back in 2008. Hopefully Braingate is robust enough to work as the connections falter.

We are still very early in the development of brain-computer interfaces, but the stakes for early adopters is very high. S3 developed tetraplegia in the late 1990s after a stroke. She, like other Braingate patients, is almost completely paralyzed. These brain implants aren’t just cool gadgets, they’re lifelines to the outside world. As this technology improves, many more people with this kind of locked-in syndrome may be able to interact with the world again by directly controlling computers with their brains. Eventually we may perfect these systems to the point where anyone can use them. When that happens, we won’t want these implants to last 1000 days, or even 10,000 days – we’ll want them to last our entire lives. After all, once you can command a machine with your thoughts, why would you ever give that up?

[image credit: Matthew McKee/BrainGate Collaboration via Brown University News]

[sources: Braingate2, Simeral et al Journal of Neural Engineering 2011 (PDF), Brown University News]