Brain-computer interfaces are opening new possibilities (2013)

nautil.us

39 points·dnetesn··5 comments

I got fairly into these years ago at the beginning of my MSc, since a research team at my Uni had cool projects about them. I gradually lost interest but I have been wondering lately if the recent advances in the Machine Learning could be capable of significantly improving the quality of some BHI application: I am happy to get pointers on papers/research labs to check out!

From what I understand, the main problem is getting the signal form individual neurons, and getting a large number of those distinct signals. I don't think there are any non-invasive methods to do that, and no one wants to do a brain surgery on a healthy brain to implant a large electrode array.

I bought an early EEG interface from MyndPlay. It still sits fairly unused in my wall-mounted cabinet. For shame.

I was pretty interested in this stuff a little while back. Daily work needs kind of steamrolled it.

I recently interviewed with another company working on these kinds of devices, however their application appears to be limited to wave tracking for meditation. They failed me for not being able to answer some vaguely-worded multithreading questions when the job was to work in web and embedded software...

I'm still intrigued. Like @winterismute said, I wonder if ML can help — maybe with pattern detection/seeing through the ..."junk".

Related open source project: http://openbci.com/

I have no affiliation with them just ran across it a while back and it has been sitting on my to-do list.

One of Elon Musk’s companies is working on this:

https://www.bloomberg.com/news/articles/2017-08-25/elon-musk...

Reply on news.ycombinator.com