Loading…

I thought non-invasive BCIs were a toy until the Cleveland Clinic thing last month | brain-computer interface - Discusd

I thought non-invasive BCIs were a toy until the Cleveland Clinic thing last month

Three weeks ago I was at a bar in Cleveland with my old lab partner from Case, and he showed me a video on his phone of a guy with a spinal injury moving his hand using a non-invasive headset, no surgery, no wires through the skull. I had spent most of grad school rolling my eyes at EEG based BCIs because the signal is noisy and the latency is garbage, and I said so. He just nodded and pulled up the data from their 2024 trial where the patient got meaningful grasp control in under 12 sessions. I went home that night and read the whole paper twice. The part that got me was they used high density dry electrodes, so no gel, and the decoder ran on a laptop, not some giant rig. I had been mentally filing non-invasive stuff under "cool demo, useless in practice" for like 8 years. Now I'm genuinely wondering if I bet on the wrong horse. Has anyone here actually tried a dry electrode setup for anything past 2D cursor control? I want to know if the fatigue issue is as bad as the old papers say.

1 comments

1 Comment

paul_thompson6710d ago

Man, I feel you on this one. I spent years telling anyone who would listen that dry electrode stuff was a dead end, and then you see a patient grabbing things after 12 sessions and it kind of messes with your head. It is not fun to admit you may have been wrong about something for almost a decade. I have not run a dry setup myself past basic cursor work, but from what I have read the fatigue thing has gotten a lot better with the newer high density arrays compared to the old gel free stuff from like 2015. Still, I would love to hear from someone who has actually put in the hours with one of these systems, because paper results and real world use are two very different things.