D
21

Kept getting garbled signals from my BCI headset for weeks

I was adjusting the electrode gel every session thinking it was a contact issue, but my buddy at the lab pointed out I had the ground reference clipped to the wrong ear clip the whole time. Switched it over and my signal clarity went from 40% to 85% accuracy on the motor imagery test. Anyone else had a setup mistake that took forever to track down?
2 comments

Log in to join the discussion

Log In
2 Comments
abby_henderson
Went from 40% to 85% accuracy" - okay, but are we sure that's really a huge deal? Like, what were you even using the headset for that 40% accuracy was so bad? I've messed up grounding clips before too and it just meant my signal was a little shaky, not totally useless. Honestly, if you're relying on a BCI for anything serious like controlling a prosthetic or something, yeah maybe that's a big jump. But for basic motor imagery tests or gaming? That's just a minor annoyance, not some crisis. I get it, it's annoying to chase a dumb mistake for weeks, but at the end of the day you found the fix and it's fine.
8
torres.jason
My lab's got a guy who spent 3 months debugging a phantom limb prosthetic on a veteran because he mounted the IMU sensor upside down on the elbow brace. The data was flipped on the y-axis. Looked like the guy was flailing when he was actually doing fine. Swapped it, 92% accuracy overnight. Ground clips are easy to miss when you're staring at software. Hardware mistakes feel personal for some reason.
3